-
-
Save stas-sl/cfa8b2dd4ca28c7382595fb86d4e9527 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
| I0330 00:58:47.679792 10583 solver.cpp:280] Solving mixed_lstm | |
| I0330 00:58:47.679805 10583 solver.cpp:281] Learning Rate Policy: fixed | |
| I0330 00:58:47.700913 10583 solver.cpp:338] Iteration 0, Testing net (#0) | |
| I0330 00:59:21.332644 10583 solver.cpp:393] Test loss: 272.275 | |
| I0330 00:59:21.333082 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0014 | |
| I0330 00:59:21.333103 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.664909 | |
| I0330 00:59:21.333117 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.00426905 | |
| I0330 00:59:21.333134 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 75.728 (* 0.3 = 22.7184 loss) | |
| I0330 00:59:21.333150 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 75.7553 (* 0.3 = 22.7266 loss) | |
| I0330 00:59:21.333165 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 00:59:21.333178 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.755637 | |
| I0330 00:59:21.333189 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.0002 | |
| I0330 00:59:21.333204 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 86.9222 (* 0.3 = 26.0767 loss) | |
| I0330 00:59:21.333219 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 86.9223 (* 0.3 = 26.0767 loss) | |
| I0330 00:59:21.333231 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 00:59:21.333243 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.75941 | |
| I0330 00:59:21.333255 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 00:59:21.333268 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 00:59:21.333283 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 00:59:21.333295 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 00:59:21.333307 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 00:59:21.642009 10583 solver.cpp:229] Iteration 0, loss = 14.0576 | |
| I0330 00:59:21.642072 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0330 00:59:21.642089 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0 | |
| I0330 00:59:21.642102 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0377358 | |
| I0330 00:59:21.642119 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.41482 (* 0.3 = 1.32445 loss) | |
| I0330 00:59:21.642133 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 4.38076 (* 0.3 = 1.31423 loss) | |
| I0330 00:59:21.642146 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0330 00:59:21.642158 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0 | |
| I0330 00:59:21.642170 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0 | |
| I0330 00:59:21.642184 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 4.36827 (* 0.3 = 1.31048 loss) | |
| I0330 00:59:21.642199 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 4.47192 (* 0.3 = 1.34157 loss) | |
| I0330 00:59:21.642210 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0 | |
| I0330 00:59:21.642222 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0 | |
| I0330 00:59:21.642233 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.0377358 | |
| I0330 00:59:21.642247 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 4.30067 (* 1 = 4.30067 loss) | |
| I0330 00:59:21.642261 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 4.46615 (* 1 = 4.46615 loss) | |
| I0330 00:59:21.642273 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 00:59:21.642284 10583 solver.cpp:245] Train net output #16: total_confidence = 4.82816e-35 | |
| I0330 00:59:21.642303 10583 sgd_solver.cpp:106] Iteration 0, lr = 0.01 | |
| I0330 00:59:21.659692 10583 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 39.1365 > 30) by scale factor 0.766548 | |
| I0330 00:59:21.967291 10583 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 48.5954 > 30) by scale factor 0.617343 | |
| I0330 00:59:22.259800 10583 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.3934 > 30) by scale factor 0.87226 | |
| I0330 00:59:22.548391 10583 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.1104 > 30) by scale factor 0.879498 | |
| I0330 00:59:22.834146 10583 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.7937 > 30) by scale factor 0.943584 | |
| I0330 01:01:41.457031 10583 solver.cpp:229] Iteration 500, loss = 8.61067 | |
| I0330 01:01:41.457216 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.047619 | |
| I0330 01:01:41.457237 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 01:01:41.457252 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.166667 | |
| I0330 01:01:41.457269 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.68505 (* 0.3 = 1.10552 loss) | |
| I0330 01:01:41.457284 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.08636 (* 0.3 = 0.325909 loss) | |
| I0330 01:01:41.457298 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.142857 | |
| I0330 01:01:41.457310 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 01:01:41.457322 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.238095 | |
| I0330 01:01:41.457336 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.62081 (* 0.3 = 1.08624 loss) | |
| I0330 01:01:41.457350 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.09615 (* 0.3 = 0.328844 loss) | |
| I0330 01:01:41.457363 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.166667 | |
| I0330 01:01:41.457376 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 01:01:41.457388 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.309524 | |
| I0330 01:01:41.457403 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.27561 (* 1 = 3.27561 loss) | |
| I0330 01:01:41.457418 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.03132 (* 1 = 1.03132 loss) | |
| I0330 01:01:41.457430 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:01:41.457443 10583 solver.cpp:245] Train net output #16: total_confidence = 9.6062e-08 | |
| I0330 01:01:41.457456 10583 sgd_solver.cpp:106] Iteration 500, lr = 0.01 | |
| I0330 01:03:59.003883 10583 solver.cpp:229] Iteration 1000, loss = 7.86221 | |
| I0330 01:03:59.004032 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0888889 | |
| I0330 01:03:59.004052 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 01:03:59.004075 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.177778 | |
| I0330 01:03:59.004091 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.86413 (* 0.3 = 1.15924 loss) | |
| I0330 01:03:59.004106 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.19489 (* 0.3 = 0.358466 loss) | |
| I0330 01:03:59.004118 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0222222 | |
| I0330 01:03:59.004132 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.698864 | |
| I0330 01:03:59.004143 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.177778 | |
| I0330 01:03:59.004158 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.73445 (* 0.3 = 1.12033 loss) | |
| I0330 01:03:59.004171 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.32688 (* 0.3 = 0.398063 loss) | |
| I0330 01:03:59.004184 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0888889 | |
| I0330 01:03:59.004195 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 01:03:59.004207 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.155556 | |
| I0330 01:03:59.004221 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.50817 (* 1 = 3.50817 loss) | |
| I0330 01:03:59.004236 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.09767 (* 1 = 1.09767 loss) | |
| I0330 01:03:59.004250 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:03:59.004261 10583 solver.cpp:245] Train net output #16: total_confidence = 4.3305e-06 | |
| I0330 01:03:59.004273 10583 sgd_solver.cpp:106] Iteration 1000, lr = 0.01 | |
| I0330 01:06:14.554929 10583 solver.cpp:229] Iteration 1500, loss = 7.63492 | |
| I0330 01:06:14.555083 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0204082 | |
| I0330 01:06:14.555104 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.704545 | |
| I0330 01:06:14.555116 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.102041 | |
| I0330 01:06:14.555133 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.27623 (* 0.3 = 1.28287 loss) | |
| I0330 01:06:14.555148 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.43714 (* 0.3 = 0.431141 loss) | |
| I0330 01:06:14.555161 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0408163 | |
| I0330 01:06:14.555173 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0330 01:06:14.555186 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.163265 | |
| I0330 01:06:14.555200 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 4.44221 (* 0.3 = 1.33266 loss) | |
| I0330 01:06:14.555214 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.38059 (* 0.3 = 0.414178 loss) | |
| I0330 01:06:14.555227 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0612245 | |
| I0330 01:06:14.555239 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.727273 | |
| I0330 01:06:14.555250 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.142857 | |
| I0330 01:06:14.555265 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 4.22528 (* 1 = 4.22528 loss) | |
| I0330 01:06:14.555279 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.26778 (* 1 = 1.26778 loss) | |
| I0330 01:06:14.555291 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:06:14.555304 10583 solver.cpp:245] Train net output #16: total_confidence = 1.00568e-05 | |
| I0330 01:06:14.555316 10583 sgd_solver.cpp:106] Iteration 1500, lr = 0.01 | |
| I0330 01:08:29.064491 10583 solver.cpp:229] Iteration 2000, loss = 7.53391 | |
| I0330 01:08:29.064647 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0363636 | |
| I0330 01:08:29.064668 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.698864 | |
| I0330 01:08:29.064682 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.127273 | |
| I0330 01:08:29.064698 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.60527 (* 0.3 = 1.08158 loss) | |
| I0330 01:08:29.064714 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.26687 (* 0.3 = 0.38006 loss) | |
| I0330 01:08:29.064728 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0545455 | |
| I0330 01:08:29.064741 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.704545 | |
| I0330 01:08:29.064754 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.254545 | |
| I0330 01:08:29.064767 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.35461 (* 0.3 = 1.00638 loss) | |
| I0330 01:08:29.064781 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.16741 (* 0.3 = 0.350224 loss) | |
| I0330 01:08:29.064793 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0727273 | |
| I0330 01:08:29.064806 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.704545 | |
| I0330 01:08:29.064818 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.272727 | |
| I0330 01:08:29.064832 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.20563 (* 1 = 3.20563 loss) | |
| I0330 01:08:29.064846 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.07118 (* 1 = 1.07118 loss) | |
| I0330 01:08:29.064858 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:08:29.064870 10583 solver.cpp:245] Train net output #16: total_confidence = 1.63657e-06 | |
| I0330 01:08:29.064883 10583 sgd_solver.cpp:106] Iteration 2000, lr = 0.01 | |
| I0330 01:10:42.997167 10583 solver.cpp:229] Iteration 2500, loss = 7.48956 | |
| I0330 01:10:42.997313 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0652174 | |
| I0330 01:10:42.997341 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 01:10:42.997354 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.130435 | |
| I0330 01:10:42.997370 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.99536 (* 0.3 = 1.19861 loss) | |
| I0330 01:10:42.997385 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.39019 (* 0.3 = 0.417056 loss) | |
| I0330 01:10:42.997397 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0217391 | |
| I0330 01:10:42.997411 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.710227 | |
| I0330 01:10:42.997422 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.108696 | |
| I0330 01:10:42.997437 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 4.09676 (* 0.3 = 1.22903 loss) | |
| I0330 01:10:42.997450 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.40789 (* 0.3 = 0.422366 loss) | |
| I0330 01:10:42.997463 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0434783 | |
| I0330 01:10:42.997475 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0330 01:10:42.997486 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.173913 | |
| I0330 01:10:42.997500 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.76722 (* 1 = 3.76722 loss) | |
| I0330 01:10:42.997514 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.12637 (* 1 = 1.12637 loss) | |
| I0330 01:10:42.997526 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:10:42.997539 10583 solver.cpp:245] Train net output #16: total_confidence = 7.18998e-06 | |
| I0330 01:10:42.997551 10583 sgd_solver.cpp:106] Iteration 2500, lr = 0.01 | |
| I0330 01:12:55.913034 10583 solver.cpp:229] Iteration 3000, loss = 7.33877 | |
| I0330 01:12:55.913182 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.04 | |
| I0330 01:12:55.913203 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0330 01:12:55.913216 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.08 | |
| I0330 01:12:55.913234 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.8715 (* 0.3 = 1.16145 loss) | |
| I0330 01:12:55.913249 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.27977 (* 0.3 = 0.383932 loss) | |
| I0330 01:12:55.913266 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.02 | |
| I0330 01:12:55.913280 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0330 01:12:55.913291 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.2 | |
| I0330 01:12:55.913306 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.70033 (* 0.3 = 1.1101 loss) | |
| I0330 01:12:55.913321 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.23724 (* 0.3 = 0.371171 loss) | |
| I0330 01:12:55.913332 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.08 | |
| I0330 01:12:55.913346 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.721591 | |
| I0330 01:12:55.913357 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.16 | |
| I0330 01:12:55.913372 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.71638 (* 1 = 3.71638 loss) | |
| I0330 01:12:55.913386 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.24898 (* 1 = 1.24898 loss) | |
| I0330 01:12:55.913398 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:12:55.913410 10583 solver.cpp:245] Train net output #16: total_confidence = 2.5845e-08 | |
| I0330 01:12:55.913424 10583 sgd_solver.cpp:106] Iteration 3000, lr = 0.01 | |
| I0330 01:15:08.318028 10583 solver.cpp:229] Iteration 3500, loss = 7.28291 | |
| I0330 01:15:08.318192 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.119048 | |
| I0330 01:15:08.318213 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 01:15:08.318227 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.166667 | |
| I0330 01:15:08.318243 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.70479 (* 0.3 = 1.11144 loss) | |
| I0330 01:15:08.318259 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.09017 (* 0.3 = 0.327051 loss) | |
| I0330 01:15:08.318271 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0952381 | |
| I0330 01:15:08.318284 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 01:15:08.318295 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.166667 | |
| I0330 01:15:08.318310 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.65345 (* 0.3 = 1.09604 loss) | |
| I0330 01:15:08.318325 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.02403 (* 0.3 = 0.307208 loss) | |
| I0330 01:15:08.318337 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0714286 | |
| I0330 01:15:08.318349 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 01:15:08.318361 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.119048 | |
| I0330 01:15:08.318377 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.65076 (* 1 = 3.65076 loss) | |
| I0330 01:15:08.318390 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.00434 (* 1 = 1.00434 loss) | |
| I0330 01:15:08.318403 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:15:08.318415 10583 solver.cpp:245] Train net output #16: total_confidence = 3.09468e-05 | |
| I0330 01:15:08.318428 10583 sgd_solver.cpp:106] Iteration 3500, lr = 0.01 | |
| I0330 01:17:20.557957 10583 solver.cpp:229] Iteration 4000, loss = 7.21408 | |
| I0330 01:17:20.558192 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0185185 | |
| I0330 01:17:20.558223 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.698864 | |
| I0330 01:17:20.558238 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0925926 | |
| I0330 01:17:20.558255 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.81017 (* 0.3 = 1.14305 loss) | |
| I0330 01:17:20.558271 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.26027 (* 0.3 = 0.378081 loss) | |
| I0330 01:17:20.558284 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.037037 | |
| I0330 01:17:20.558297 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.6875 | |
| I0330 01:17:20.558310 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.148148 | |
| I0330 01:17:20.558323 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.82885 (* 0.3 = 1.14865 loss) | |
| I0330 01:17:20.558338 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.31299 (* 0.3 = 0.393896 loss) | |
| I0330 01:17:20.558351 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.037037 | |
| I0330 01:17:20.558363 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.704545 | |
| I0330 01:17:20.558377 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.148148 | |
| I0330 01:17:20.558390 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.69307 (* 1 = 3.69307 loss) | |
| I0330 01:17:20.558405 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.18354 (* 1 = 1.18354 loss) | |
| I0330 01:17:20.558418 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:17:20.558429 10583 solver.cpp:245] Train net output #16: total_confidence = 1.23132e-06 | |
| I0330 01:17:20.558444 10583 sgd_solver.cpp:106] Iteration 4000, lr = 0.01 | |
| I0330 01:19:33.516381 10583 solver.cpp:229] Iteration 4500, loss = 7.12986 | |
| I0330 01:19:33.516528 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0392157 | |
| I0330 01:19:33.516548 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.676136 | |
| I0330 01:19:33.516562 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.196078 | |
| I0330 01:19:33.516578 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.78394 (* 0.3 = 1.13518 loss) | |
| I0330 01:19:33.516593 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.40831 (* 0.3 = 0.422492 loss) | |
| I0330 01:19:33.516605 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0588235 | |
| I0330 01:19:33.516618 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0330 01:19:33.516630 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.235294 | |
| I0330 01:19:33.516644 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.84204 (* 0.3 = 1.15261 loss) | |
| I0330 01:19:33.516659 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.21641 (* 0.3 = 0.364922 loss) | |
| I0330 01:19:33.516670 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0588235 | |
| I0330 01:19:33.516682 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.721591 | |
| I0330 01:19:33.516695 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.235294 | |
| I0330 01:19:33.516710 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.57727 (* 1 = 3.57727 loss) | |
| I0330 01:19:33.516722 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.12683 (* 1 = 1.12683 loss) | |
| I0330 01:19:33.516736 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:19:33.516747 10583 solver.cpp:245] Train net output #16: total_confidence = 2.97965e-05 | |
| I0330 01:19:33.516760 10583 sgd_solver.cpp:106] Iteration 4500, lr = 0.01 | |
| I0330 01:21:45.840735 10583 solver.cpp:338] Iteration 5000, Testing net (#0) | |
| I0330 01:22:16.200706 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 01:22:16.200825 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 01:22:16.200846 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760319 | |
| I0330 01:22:16.200860 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 01:22:16.200876 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:22:16.200893 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:22:16.200906 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 01:22:16.200917 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760319 | |
| I0330 01:22:16.200932 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 01:22:16.200945 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:22:16.200960 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:22:16.200973 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 01:22:16.200984 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760319 | |
| I0330 01:22:16.200996 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 01:22:16.201010 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 01:22:16.201025 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 01:22:16.201037 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 01:22:16.201050 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 01:22:16.353761 10583 solver.cpp:229] Iteration 5000, loss = 7.13386 | |
| I0330 01:22:16.353857 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0217391 | |
| I0330 01:22:16.353880 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0330 01:22:16.353894 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.152174 | |
| I0330 01:22:16.353914 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.3741 (* 0.3 = 1.01223 loss) | |
| I0330 01:22:16.353929 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.12248 (* 0.3 = 0.336744 loss) | |
| I0330 01:22:16.353941 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0869565 | |
| I0330 01:22:16.353955 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 01:22:16.353967 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.195652 | |
| I0330 01:22:16.353981 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.31311 (* 0.3 = 0.993933 loss) | |
| I0330 01:22:16.353996 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.980257 (* 0.3 = 0.294077 loss) | |
| I0330 01:22:16.354008 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0652174 | |
| I0330 01:22:16.354022 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 01:22:16.354034 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.195652 | |
| I0330 01:22:16.354048 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.19737 (* 1 = 3.19737 loss) | |
| I0330 01:22:16.354063 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.943961 (* 1 = 0.943961 loss) | |
| I0330 01:22:16.354075 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:22:16.354089 10583 solver.cpp:245] Train net output #16: total_confidence = 6.88434e-05 | |
| I0330 01:22:16.354102 10583 sgd_solver.cpp:106] Iteration 5000, lr = 0.01 | |
| I0330 01:24:28.277389 10583 solver.cpp:229] Iteration 5500, loss = 7.03053 | |
| I0330 01:24:28.277601 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0444444 | |
| I0330 01:24:28.277622 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 01:24:28.277644 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0888889 | |
| I0330 01:24:28.277662 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.22592 (* 0.3 = 0.967777 loss) | |
| I0330 01:24:28.277678 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02526 (* 0.3 = 0.307579 loss) | |
| I0330 01:24:28.277690 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0222222 | |
| I0330 01:24:28.277704 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 01:24:28.277715 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.222222 | |
| I0330 01:24:28.277729 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.34224 (* 0.3 = 1.00267 loss) | |
| I0330 01:24:28.277745 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.971351 (* 0.3 = 0.291405 loss) | |
| I0330 01:24:28.277757 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0222222 | |
| I0330 01:24:28.277770 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0330 01:24:28.277782 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.2 | |
| I0330 01:24:28.277797 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.12382 (* 1 = 3.12382 loss) | |
| I0330 01:24:28.277812 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.897123 (* 1 = 0.897123 loss) | |
| I0330 01:24:28.277824 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:24:28.277837 10583 solver.cpp:245] Train net output #16: total_confidence = 1.74858e-06 | |
| I0330 01:24:28.277849 10583 sgd_solver.cpp:106] Iteration 5500, lr = 0.01 | |
| I0330 01:26:40.308495 10583 solver.cpp:229] Iteration 6000, loss = 7.04775 | |
| I0330 01:26:40.308678 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0263158 | |
| I0330 01:26:40.308701 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 01:26:40.308714 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.105263 | |
| I0330 01:26:40.308732 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.6265 (* 0.3 = 1.08795 loss) | |
| I0330 01:26:40.308748 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.06268 (* 0.3 = 0.318804 loss) | |
| I0330 01:26:40.308760 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0263158 | |
| I0330 01:26:40.308773 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 01:26:40.308785 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.157895 | |
| I0330 01:26:40.308799 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.66521 (* 0.3 = 1.09956 loss) | |
| I0330 01:26:40.308815 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.943268 (* 0.3 = 0.28298 loss) | |
| I0330 01:26:40.308826 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0526316 | |
| I0330 01:26:40.308840 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0330 01:26:40.308851 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.157895 | |
| I0330 01:26:40.308866 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.50392 (* 1 = 3.50392 loss) | |
| I0330 01:26:40.308881 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.926607 (* 1 = 0.926607 loss) | |
| I0330 01:26:40.308893 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:26:40.308905 10583 solver.cpp:245] Train net output #16: total_confidence = 9.23597e-05 | |
| I0330 01:26:40.308919 10583 sgd_solver.cpp:106] Iteration 6000, lr = 0.01 | |
| I0330 01:28:52.020663 10583 solver.cpp:229] Iteration 6500, loss = 6.98487 | |
| I0330 01:28:52.020885 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0330 01:28:52.020906 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.698864 | |
| I0330 01:28:52.020930 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.132075 | |
| I0330 01:28:52.020946 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.48622 (* 0.3 = 1.04587 loss) | |
| I0330 01:28:52.020961 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.14289 (* 0.3 = 0.342866 loss) | |
| I0330 01:28:52.020974 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0566038 | |
| I0330 01:28:52.020987 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0330 01:28:52.020999 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.150943 | |
| I0330 01:28:52.021014 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.60393 (* 0.3 = 1.08118 loss) | |
| I0330 01:28:52.021028 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.13711 (* 0.3 = 0.341132 loss) | |
| I0330 01:28:52.021040 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0188679 | |
| I0330 01:28:52.021054 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.704545 | |
| I0330 01:28:52.021065 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.226415 | |
| I0330 01:28:52.021080 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.38148 (* 1 = 3.38148 loss) | |
| I0330 01:28:52.021095 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.06576 (* 1 = 1.06576 loss) | |
| I0330 01:28:52.021106 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:28:52.021118 10583 solver.cpp:245] Train net output #16: total_confidence = 6.25799e-07 | |
| I0330 01:28:52.021131 10583 sgd_solver.cpp:106] Iteration 6500, lr = 0.01 | |
| I0330 01:31:03.487447 10583 solver.cpp:229] Iteration 7000, loss = 6.90967 | |
| I0330 01:31:03.487602 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0444444 | |
| I0330 01:31:03.487622 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 01:31:03.487637 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.177778 | |
| I0330 01:31:03.487653 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.11284 (* 0.3 = 0.933853 loss) | |
| I0330 01:31:03.487668 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.95473 (* 0.3 = 0.286419 loss) | |
| I0330 01:31:03.487679 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0888889 | |
| I0330 01:31:03.487692 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 01:31:03.487704 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.266667 | |
| I0330 01:31:03.487718 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.069 (* 0.3 = 0.920699 loss) | |
| I0330 01:31:03.487732 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.00588 (* 0.3 = 0.301764 loss) | |
| I0330 01:31:03.487746 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0888889 | |
| I0330 01:31:03.487757 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 01:31:03.487769 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.244444 | |
| I0330 01:31:03.487784 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.02472 (* 1 = 3.02472 loss) | |
| I0330 01:31:03.487798 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.851511 (* 1 = 0.851511 loss) | |
| I0330 01:31:03.487810 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:31:03.487823 10583 solver.cpp:245] Train net output #16: total_confidence = 6.78319e-06 | |
| I0330 01:31:03.487835 10583 sgd_solver.cpp:106] Iteration 7000, lr = 0.01 | |
| I0330 01:33:15.011142 10583 solver.cpp:229] Iteration 7500, loss = 6.88338 | |
| I0330 01:33:15.011368 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0392157 | |
| I0330 01:33:15.011389 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.710227 | |
| I0330 01:33:15.011404 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.176471 | |
| I0330 01:33:15.011421 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.3905 (* 0.3 = 1.01715 loss) | |
| I0330 01:33:15.011436 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.07501 (* 0.3 = 0.322503 loss) | |
| I0330 01:33:15.011456 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0588235 | |
| I0330 01:33:15.011468 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0330 01:33:15.011481 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.196078 | |
| I0330 01:33:15.011495 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.49042 (* 0.3 = 1.04713 loss) | |
| I0330 01:33:15.011518 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.17853 (* 0.3 = 0.35356 loss) | |
| I0330 01:33:15.011531 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0588235 | |
| I0330 01:33:15.011544 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.721591 | |
| I0330 01:33:15.011556 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.196078 | |
| I0330 01:33:15.011570 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.32681 (* 1 = 3.32681 loss) | |
| I0330 01:33:15.011585 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.01018 (* 1 = 1.01018 loss) | |
| I0330 01:33:15.011597 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:33:15.011610 10583 solver.cpp:245] Train net output #16: total_confidence = 5.73006e-05 | |
| I0330 01:33:15.011623 10583 sgd_solver.cpp:106] Iteration 7500, lr = 0.01 | |
| I0330 01:35:26.089201 10583 solver.cpp:229] Iteration 8000, loss = 6.90374 | |
| I0330 01:35:26.089349 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0526316 | |
| I0330 01:35:26.089371 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.6875 | |
| I0330 01:35:26.089385 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.157895 | |
| I0330 01:35:26.089401 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.44948 (* 0.3 = 1.03484 loss) | |
| I0330 01:35:26.089416 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.20612 (* 0.3 = 0.361836 loss) | |
| I0330 01:35:26.089429 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.122807 | |
| I0330 01:35:26.089442 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0330 01:35:26.089454 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.157895 | |
| I0330 01:35:26.089469 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.43902 (* 0.3 = 1.03171 loss) | |
| I0330 01:35:26.089483 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.16705 (* 0.3 = 0.350115 loss) | |
| I0330 01:35:26.089495 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0350877 | |
| I0330 01:35:26.089510 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.6875 | |
| I0330 01:35:26.089532 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.192982 | |
| I0330 01:35:26.089560 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.26792 (* 1 = 3.26792 loss) | |
| I0330 01:35:26.089577 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.0926 (* 1 = 1.0926 loss) | |
| I0330 01:35:26.089591 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:35:26.089602 10583 solver.cpp:245] Train net output #16: total_confidence = 9.21544e-06 | |
| I0330 01:35:26.089617 10583 sgd_solver.cpp:106] Iteration 8000, lr = 0.01 | |
| I0330 01:37:37.386723 10583 solver.cpp:229] Iteration 8500, loss = 6.89902 | |
| I0330 01:37:37.386930 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0434783 | |
| I0330 01:37:37.386952 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 01:37:37.386966 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.108696 | |
| I0330 01:37:37.386986 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.67592 (* 0.3 = 1.10278 loss) | |
| I0330 01:37:37.387001 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.04447 (* 0.3 = 0.313342 loss) | |
| I0330 01:37:37.387013 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.108696 | |
| I0330 01:37:37.387027 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 01:37:37.387039 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.23913 | |
| I0330 01:37:37.387068 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.5104 (* 0.3 = 1.05312 loss) | |
| I0330 01:37:37.387082 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.993681 (* 0.3 = 0.298104 loss) | |
| I0330 01:37:37.387095 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0869565 | |
| I0330 01:37:37.387109 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0330 01:37:37.387120 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.195652 | |
| I0330 01:37:37.387135 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.33472 (* 1 = 3.33472 loss) | |
| I0330 01:37:37.387150 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.95642 (* 1 = 0.95642 loss) | |
| I0330 01:37:37.387163 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:37:37.387176 10583 solver.cpp:245] Train net output #16: total_confidence = 3.09936e-05 | |
| I0330 01:37:37.387190 10583 sgd_solver.cpp:106] Iteration 8500, lr = 0.01 | |
| I0330 01:39:48.628504 10583 solver.cpp:229] Iteration 9000, loss = 6.82579 | |
| I0330 01:39:48.628655 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.06 | |
| I0330 01:39:48.628685 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0330 01:39:48.628710 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.18 | |
| I0330 01:39:48.628726 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.73238 (* 0.3 = 1.11971 loss) | |
| I0330 01:39:48.628741 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.15983 (* 0.3 = 0.347949 loss) | |
| I0330 01:39:48.628754 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.04 | |
| I0330 01:39:48.628767 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.693182 | |
| I0330 01:39:48.628778 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.1 | |
| I0330 01:39:48.628793 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.72515 (* 0.3 = 1.11755 loss) | |
| I0330 01:39:48.628808 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.25802 (* 0.3 = 0.377407 loss) | |
| I0330 01:39:48.628820 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.06 | |
| I0330 01:39:48.628832 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.715909 | |
| I0330 01:39:48.628844 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.14 | |
| I0330 01:39:48.628859 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.44993 (* 1 = 3.44993 loss) | |
| I0330 01:39:48.628873 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.0567 (* 1 = 1.0567 loss) | |
| I0330 01:39:48.628885 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:39:48.628897 10583 solver.cpp:245] Train net output #16: total_confidence = 2.08854e-06 | |
| I0330 01:39:48.628911 10583 sgd_solver.cpp:106] Iteration 9000, lr = 0.01 | |
| I0330 01:41:59.844422 10583 solver.cpp:229] Iteration 9500, loss = 6.83051 | |
| I0330 01:41:59.844626 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0625 | |
| I0330 01:41:59.844647 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 01:41:59.844661 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.25 | |
| I0330 01:41:59.844678 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.59601 (* 0.3 = 1.0788 loss) | |
| I0330 01:41:59.844693 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.0404 (* 0.3 = 0.312121 loss) | |
| I0330 01:41:59.844707 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.104167 | |
| I0330 01:41:59.844719 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 01:41:59.844732 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.166667 | |
| I0330 01:41:59.844746 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.60519 (* 0.3 = 1.08156 loss) | |
| I0330 01:41:59.844761 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.05906 (* 0.3 = 0.317718 loss) | |
| I0330 01:41:59.844774 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0833333 | |
| I0330 01:41:59.844785 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0330 01:41:59.844797 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.125 | |
| I0330 01:41:59.844811 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.45664 (* 1 = 3.45664 loss) | |
| I0330 01:41:59.844825 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.984761 (* 1 = 0.984761 loss) | |
| I0330 01:41:59.844837 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:41:59.844849 10583 solver.cpp:245] Train net output #16: total_confidence = 1.07748e-05 | |
| I0330 01:41:59.844863 10583 sgd_solver.cpp:106] Iteration 9500, lr = 0.01 | |
| I0330 01:44:11.031463 10583 solver.cpp:338] Iteration 10000, Testing net (#0) | |
| I0330 01:44:41.364166 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 01:44:41.364310 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 01:44:41.364329 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759228 | |
| I0330 01:44:41.364343 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 01:44:41.364361 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:44:41.364377 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:44:41.364389 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 01:44:41.364401 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759228 | |
| I0330 01:44:41.364413 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 01:44:41.364428 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:44:41.364442 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 01:44:41.364455 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 01:44:41.364467 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759228 | |
| I0330 01:44:41.364478 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 01:44:41.364493 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 01:44:41.364508 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 01:44:41.364521 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 01:44:41.364532 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 01:44:41.516947 10583 solver.cpp:229] Iteration 10000, loss = 6.80429 | |
| I0330 01:44:41.517021 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0425532 | |
| I0330 01:44:41.517040 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 01:44:41.517055 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.170213 | |
| I0330 01:44:41.517071 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.57653 (* 0.3 = 1.07296 loss) | |
| I0330 01:44:41.517087 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.09159 (* 0.3 = 0.327477 loss) | |
| I0330 01:44:41.517099 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0851064 | |
| I0330 01:44:41.517112 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 01:44:41.517124 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.212766 | |
| I0330 01:44:41.517139 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.38877 (* 0.3 = 1.01663 loss) | |
| I0330 01:44:41.517153 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.10335 (* 0.3 = 0.331006 loss) | |
| I0330 01:44:41.517165 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0851064 | |
| I0330 01:44:41.517179 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0330 01:44:41.517190 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.255319 | |
| I0330 01:44:41.517204 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.2213 (* 1 = 3.2213 loss) | |
| I0330 01:44:41.517220 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.01048 (* 1 = 1.01048 loss) | |
| I0330 01:44:41.517232 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:44:41.517244 10583 solver.cpp:245] Train net output #16: total_confidence = 6.35376e-06 | |
| I0330 01:44:41.517258 10583 sgd_solver.cpp:106] Iteration 10000, lr = 0.01 | |
| I0330 01:46:52.725340 10583 solver.cpp:229] Iteration 10500, loss = 6.76326 | |
| I0330 01:46:52.725513 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535 | |
| I0330 01:46:52.725534 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 01:46:52.725548 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.209302 | |
| I0330 01:46:52.725564 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.2325 (* 0.3 = 0.969749 loss) | |
| I0330 01:46:52.725580 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.940328 (* 0.3 = 0.282098 loss) | |
| I0330 01:46:52.725594 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0697674 | |
| I0330 01:46:52.725605 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0330 01:46:52.725618 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.255814 | |
| I0330 01:46:52.725633 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.30093 (* 0.3 = 0.99028 loss) | |
| I0330 01:46:52.725647 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.990723 (* 0.3 = 0.297217 loss) | |
| I0330 01:46:52.725659 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.116279 | |
| I0330 01:46:52.725672 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 01:46:52.725684 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.209302 | |
| I0330 01:46:52.725698 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.21785 (* 1 = 3.21785 loss) | |
| I0330 01:46:52.725713 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.910933 (* 1 = 0.910933 loss) | |
| I0330 01:46:52.725725 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:46:52.725738 10583 solver.cpp:245] Train net output #16: total_confidence = 4.19291e-05 | |
| I0330 01:46:52.725754 10583 sgd_solver.cpp:106] Iteration 10500, lr = 0.01 | |
| I0330 01:49:03.657474 10583 solver.cpp:229] Iteration 11000, loss = 6.73095 | |
| I0330 01:49:03.657665 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.125 | |
| I0330 01:49:03.657694 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 01:49:03.657707 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.291667 | |
| I0330 01:49:03.657724 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.17677 (* 0.3 = 0.953031 loss) | |
| I0330 01:49:03.657740 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.97894 (* 0.3 = 0.293682 loss) | |
| I0330 01:49:03.657753 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.145833 | |
| I0330 01:49:03.657765 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 01:49:03.657778 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.3125 | |
| I0330 01:49:03.657796 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.18182 (* 0.3 = 0.954545 loss) | |
| I0330 01:49:03.657810 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.933686 (* 0.3 = 0.280106 loss) | |
| I0330 01:49:03.657824 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.125 | |
| I0330 01:49:03.657835 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 01:49:03.657856 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.333333 | |
| I0330 01:49:03.657871 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.14034 (* 1 = 3.14034 loss) | |
| I0330 01:49:03.657886 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.922778 (* 1 = 0.922778 loss) | |
| I0330 01:49:03.657897 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:49:03.657910 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000118025 | |
| I0330 01:49:03.657924 10583 sgd_solver.cpp:106] Iteration 11000, lr = 0.01 | |
| I0330 01:51:13.975004 10583 solver.cpp:229] Iteration 11500, loss = 6.72569 | |
| I0330 01:51:13.975147 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0816327 | |
| I0330 01:51:13.975178 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 01:51:13.975191 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.163265 | |
| I0330 01:51:13.975208 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.71438 (* 0.3 = 1.11431 loss) | |
| I0330 01:51:13.975224 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.11374 (* 0.3 = 0.334122 loss) | |
| I0330 01:51:13.975235 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0612245 | |
| I0330 01:51:13.975249 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0330 01:51:13.975260 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.265306 | |
| I0330 01:51:13.975275 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.58644 (* 0.3 = 1.07593 loss) | |
| I0330 01:51:13.975289 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.10777 (* 0.3 = 0.332332 loss) | |
| I0330 01:51:13.975301 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.102041 | |
| I0330 01:51:13.975314 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.732955 | |
| I0330 01:51:13.975327 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.22449 | |
| I0330 01:51:13.975340 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.4235 (* 1 = 3.4235 loss) | |
| I0330 01:51:13.975354 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.09914 (* 1 = 1.09914 loss) | |
| I0330 01:51:13.975366 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:51:13.975378 10583 solver.cpp:245] Train net output #16: total_confidence = 5.82417e-06 | |
| I0330 01:51:13.975391 10583 sgd_solver.cpp:106] Iteration 11500, lr = 0.01 | |
| I0330 01:53:23.892590 10583 solver.cpp:229] Iteration 12000, loss = 6.65187 | |
| I0330 01:53:23.892737 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.205128 | |
| I0330 01:53:23.892758 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 01:53:23.892771 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.333333 | |
| I0330 01:53:23.892788 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03764 (* 0.3 = 0.911293 loss) | |
| I0330 01:53:23.892802 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.826328 (* 0.3 = 0.247898 loss) | |
| I0330 01:53:23.892814 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.179487 | |
| I0330 01:53:23.892827 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 01:53:23.892838 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.384615 | |
| I0330 01:53:23.892853 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.9077 (* 0.3 = 0.872311 loss) | |
| I0330 01:53:23.892866 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.822541 (* 0.3 = 0.246762 loss) | |
| I0330 01:53:23.892879 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.179487 | |
| I0330 01:53:23.892890 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0330 01:53:23.892902 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.307692 | |
| I0330 01:53:23.892916 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.85254 (* 1 = 2.85254 loss) | |
| I0330 01:53:23.892930 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.74541 (* 1 = 0.74541 loss) | |
| I0330 01:53:23.892942 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:53:23.892953 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000715381 | |
| I0330 01:53:23.892966 10583 sgd_solver.cpp:106] Iteration 12000, lr = 0.01 | |
| I0330 01:55:34.311179 10583 solver.cpp:229] Iteration 12500, loss = 6.67389 | |
| I0330 01:55:34.311306 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0851064 | |
| I0330 01:55:34.311326 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 01:55:34.311339 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.234043 | |
| I0330 01:55:34.311355 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.43327 (* 0.3 = 1.02998 loss) | |
| I0330 01:55:34.311370 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.054 (* 0.3 = 0.3162 loss) | |
| I0330 01:55:34.311383 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0851064 | |
| I0330 01:55:34.311395 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 01:55:34.311408 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.170213 | |
| I0330 01:55:34.311421 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.38029 (* 0.3 = 1.01409 loss) | |
| I0330 01:55:34.311437 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.03973 (* 0.3 = 0.311919 loss) | |
| I0330 01:55:34.311450 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.106383 | |
| I0330 01:55:34.311461 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0330 01:55:34.311473 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.234043 | |
| I0330 01:55:34.311487 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.23909 (* 1 = 3.23909 loss) | |
| I0330 01:55:34.311501 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.919139 (* 1 = 0.919139 loss) | |
| I0330 01:55:34.311513 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:55:34.311525 10583 solver.cpp:245] Train net output #16: total_confidence = 5.49713e-06 | |
| I0330 01:55:34.311537 10583 sgd_solver.cpp:106] Iteration 12500, lr = 0.01 | |
| I0330 01:57:44.054023 10583 solver.cpp:229] Iteration 13000, loss = 6.63888 | |
| I0330 01:57:44.054165 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.146341 | |
| I0330 01:57:44.054185 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 01:57:44.054199 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.195122 | |
| I0330 01:57:44.054215 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.39822 (* 0.3 = 1.01947 loss) | |
| I0330 01:57:44.054230 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.957893 (* 0.3 = 0.287368 loss) | |
| I0330 01:57:44.054244 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.097561 | |
| I0330 01:57:44.054256 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 01:57:44.054268 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.195122 | |
| I0330 01:57:44.054283 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.40589 (* 0.3 = 1.02177 loss) | |
| I0330 01:57:44.054297 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.932125 (* 0.3 = 0.279638 loss) | |
| I0330 01:57:44.054311 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.146341 | |
| I0330 01:57:44.054323 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 01:57:44.054335 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.243902 | |
| I0330 01:57:44.054349 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.31721 (* 1 = 3.31721 loss) | |
| I0330 01:57:44.054364 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.877534 (* 1 = 0.877534 loss) | |
| I0330 01:57:44.054378 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:57:44.054388 10583 solver.cpp:245] Train net output #16: total_confidence = 3.52872e-05 | |
| I0330 01:57:44.054402 10583 sgd_solver.cpp:106] Iteration 13000, lr = 0.01 | |
| I0330 01:59:53.829144 10583 solver.cpp:229] Iteration 13500, loss = 6.61232 | |
| I0330 01:59:53.829260 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.106383 | |
| I0330 01:59:53.829279 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 01:59:53.829293 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.234043 | |
| I0330 01:59:53.829309 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.58177 (* 0.3 = 1.07453 loss) | |
| I0330 01:59:53.829324 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.16729 (* 0.3 = 0.350186 loss) | |
| I0330 01:59:53.829336 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0851064 | |
| I0330 01:59:53.829349 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0330 01:59:53.829360 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.212766 | |
| I0330 01:59:53.829375 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.66689 (* 0.3 = 1.10007 loss) | |
| I0330 01:59:53.829388 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.22142 (* 0.3 = 0.366426 loss) | |
| I0330 01:59:53.829401 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.148936 | |
| I0330 01:59:53.829412 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 01:59:53.829424 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.319149 | |
| I0330 01:59:53.829438 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.61029 (* 1 = 3.61029 loss) | |
| I0330 01:59:53.829452 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.14352 (* 1 = 1.14352 loss) | |
| I0330 01:59:53.829464 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 01:59:53.829476 10583 solver.cpp:245] Train net output #16: total_confidence = 4.52149e-05 | |
| I0330 01:59:53.829488 10583 sgd_solver.cpp:106] Iteration 13500, lr = 0.01 | |
| I0330 02:02:03.753356 10583 solver.cpp:229] Iteration 14000, loss = 6.59003 | |
| I0330 02:02:03.753496 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.116279 | |
| I0330 02:02:03.753517 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 02:02:03.753530 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.232558 | |
| I0330 02:02:03.753548 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.27967 (* 0.3 = 0.983902 loss) | |
| I0330 02:02:03.753563 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.881889 (* 0.3 = 0.264567 loss) | |
| I0330 02:02:03.753576 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0465116 | |
| I0330 02:02:03.753588 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0330 02:02:03.753602 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.186047 | |
| I0330 02:02:03.753615 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.28546 (* 0.3 = 0.985637 loss) | |
| I0330 02:02:03.753629 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.93741 (* 0.3 = 0.281223 loss) | |
| I0330 02:02:03.753643 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.116279 | |
| I0330 02:02:03.753654 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0330 02:02:03.753666 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.302326 | |
| I0330 02:02:03.753680 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.14972 (* 1 = 3.14972 loss) | |
| I0330 02:02:03.753695 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.864224 (* 1 = 0.864224 loss) | |
| I0330 02:02:03.753707 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:02:03.753720 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000228823 | |
| I0330 02:02:03.753732 10583 sgd_solver.cpp:106] Iteration 14000, lr = 0.01 | |
| I0330 02:04:13.581449 10583 solver.cpp:229] Iteration 14500, loss = 6.56195 | |
| I0330 02:04:13.581568 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.06 | |
| I0330 02:04:13.581588 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 02:04:13.581600 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.12 | |
| I0330 02:04:13.581617 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.41784 (* 0.3 = 1.02535 loss) | |
| I0330 02:04:13.581632 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.01301 (* 0.3 = 0.303904 loss) | |
| I0330 02:04:13.581645 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.02 | |
| I0330 02:04:13.581657 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0330 02:04:13.581670 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.16 | |
| I0330 02:04:13.581684 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.37966 (* 0.3 = 1.0139 loss) | |
| I0330 02:04:13.581699 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.00877 (* 0.3 = 0.302632 loss) | |
| I0330 02:04:13.581712 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.06 | |
| I0330 02:04:13.581723 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.732955 | |
| I0330 02:04:13.581735 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.18 | |
| I0330 02:04:13.581749 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.3567 (* 1 = 3.3567 loss) | |
| I0330 02:04:13.581764 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.989174 (* 1 = 0.989174 loss) | |
| I0330 02:04:13.581776 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:04:13.581789 10583 solver.cpp:245] Train net output #16: total_confidence = 2.71919e-06 | |
| I0330 02:04:13.581801 10583 sgd_solver.cpp:106] Iteration 14500, lr = 0.01 | |
| I0330 02:06:23.158325 10583 solver.cpp:338] Iteration 15000, Testing net (#0) | |
| I0330 02:06:52.841222 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 02:06:52.841269 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 02:06:52.841285 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759728 | |
| I0330 02:06:52.841300 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 02:06:52.841315 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:06:52.841331 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:06:52.841344 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 02:06:52.841356 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759728 | |
| I0330 02:06:52.841367 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 02:06:52.841382 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:06:52.841397 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:06:52.841408 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 02:06:52.841421 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759728 | |
| I0330 02:06:52.841434 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 02:06:52.841447 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:06:52.841461 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:06:52.841475 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 02:06:52.841485 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 02:06:52.992192 10583 solver.cpp:229] Iteration 15000, loss = 6.57188 | |
| I0330 02:06:52.992231 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0980392 | |
| I0330 02:06:52.992249 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 02:06:52.992262 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.294118 | |
| I0330 02:06:52.992279 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.11547 (* 0.3 = 0.934642 loss) | |
| I0330 02:06:52.992293 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.00679 (* 0.3 = 0.302036 loss) | |
| I0330 02:06:52.992305 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0588235 | |
| I0330 02:06:52.992318 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.710227 | |
| I0330 02:06:52.992331 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.176471 | |
| I0330 02:06:52.992347 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.0998 (* 0.3 = 0.92994 loss) | |
| I0330 02:06:52.992360 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.01195 (* 0.3 = 0.303586 loss) | |
| I0330 02:06:52.992373 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0980392 | |
| I0330 02:06:52.992385 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0330 02:06:52.992398 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.294118 | |
| I0330 02:06:52.992413 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.92032 (* 1 = 2.92032 loss) | |
| I0330 02:06:52.992426 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.916056 (* 1 = 0.916056 loss) | |
| I0330 02:06:52.992439 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:06:52.992450 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0001723 | |
| I0330 02:06:52.992463 10583 sgd_solver.cpp:106] Iteration 15000, lr = 0.01 | |
| I0330 02:09:02.375906 10583 solver.cpp:229] Iteration 15500, loss = 6.5487 | |
| I0330 02:09:02.376052 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0625 | |
| I0330 02:09:02.376072 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 02:09:02.376085 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.208333 | |
| I0330 02:09:02.376103 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.20917 (* 0.3 = 0.962752 loss) | |
| I0330 02:09:02.376117 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.932314 (* 0.3 = 0.279694 loss) | |
| I0330 02:09:02.376130 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0416667 | |
| I0330 02:09:02.376142 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0330 02:09:02.376154 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.166667 | |
| I0330 02:09:02.376171 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.25082 (* 0.3 = 0.975247 loss) | |
| I0330 02:09:02.376186 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.93558 (* 0.3 = 0.280674 loss) | |
| I0330 02:09:02.376199 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.104167 | |
| I0330 02:09:02.376210 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0330 02:09:02.376222 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.229167 | |
| I0330 02:09:02.376237 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.12187 (* 1 = 3.12187 loss) | |
| I0330 02:09:02.376251 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.883792 (* 1 = 0.883792 loss) | |
| I0330 02:09:02.376263 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:09:02.376276 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000894235 | |
| I0330 02:09:02.376287 10583 sgd_solver.cpp:106] Iteration 15500, lr = 0.01 | |
| I0330 02:11:11.903537 10583 solver.cpp:229] Iteration 16000, loss = 6.50345 | |
| I0330 02:11:11.903898 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0612245 | |
| I0330 02:11:11.903918 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 02:11:11.903931 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.183673 | |
| I0330 02:11:11.903949 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.64614 (* 0.3 = 1.09384 loss) | |
| I0330 02:11:11.903964 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.07499 (* 0.3 = 0.322498 loss) | |
| I0330 02:11:11.903975 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0816327 | |
| I0330 02:11:11.903988 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 02:11:11.904000 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.163265 | |
| I0330 02:11:11.904013 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.60183 (* 0.3 = 1.08055 loss) | |
| I0330 02:11:11.904028 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.06643 (* 0.3 = 0.31993 loss) | |
| I0330 02:11:11.904041 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0612245 | |
| I0330 02:11:11.904052 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0330 02:11:11.904064 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.142857 | |
| I0330 02:11:11.904078 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.45324 (* 1 = 3.45324 loss) | |
| I0330 02:11:11.904093 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.02502 (* 1 = 1.02502 loss) | |
| I0330 02:11:11.904105 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:11:11.904117 10583 solver.cpp:245] Train net output #16: total_confidence = 5.59599e-05 | |
| I0330 02:11:11.904129 10583 sgd_solver.cpp:106] Iteration 16000, lr = 0.01 | |
| I0330 02:13:22.062325 10583 solver.cpp:229] Iteration 16500, loss = 6.54353 | |
| I0330 02:13:22.062515 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0434783 | |
| I0330 02:13:22.062536 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 02:13:22.062551 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.26087 | |
| I0330 02:13:22.062567 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03688 (* 0.3 = 0.911063 loss) | |
| I0330 02:13:22.062582 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.822983 (* 0.3 = 0.246895 loss) | |
| I0330 02:13:22.062595 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0434783 | |
| I0330 02:13:22.062608 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 02:13:22.062620 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.173913 | |
| I0330 02:13:22.062634 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.04935 (* 0.3 = 0.914805 loss) | |
| I0330 02:13:22.062649 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.836682 (* 0.3 = 0.251005 loss) | |
| I0330 02:13:22.062661 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0869565 | |
| I0330 02:13:22.062674 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0330 02:13:22.062685 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.304348 | |
| I0330 02:13:22.062700 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9387 (* 1 = 2.9387 loss) | |
| I0330 02:13:22.062712 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.833139 (* 1 = 0.833139 loss) | |
| I0330 02:13:22.062726 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:13:22.062736 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000136418 | |
| I0330 02:13:22.062750 10583 sgd_solver.cpp:106] Iteration 16500, lr = 0.01 | |
| I0330 02:15:31.341727 10583 solver.cpp:229] Iteration 17000, loss = 6.46967 | |
| I0330 02:15:31.341843 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.162791 | |
| I0330 02:15:31.341862 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 02:15:31.341876 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.232558 | |
| I0330 02:15:31.341892 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.07968 (* 0.3 = 0.923903 loss) | |
| I0330 02:15:31.341907 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.806703 (* 0.3 = 0.242011 loss) | |
| I0330 02:15:31.341919 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.139535 | |
| I0330 02:15:31.341933 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 02:15:31.341944 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.27907 | |
| I0330 02:15:31.341958 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.00357 (* 0.3 = 0.90107 loss) | |
| I0330 02:15:31.341972 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.787483 (* 0.3 = 0.236245 loss) | |
| I0330 02:15:31.341985 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.116279 | |
| I0330 02:15:31.341997 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 02:15:31.342010 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.232558 | |
| I0330 02:15:31.342023 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.8475 (* 1 = 2.8475 loss) | |
| I0330 02:15:31.342037 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.769009 (* 1 = 0.769009 loss) | |
| I0330 02:15:31.342049 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:15:31.342061 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000247626 | |
| I0330 02:15:31.342074 10583 sgd_solver.cpp:106] Iteration 17000, lr = 0.01 | |
| I0330 02:17:40.990178 10583 solver.cpp:229] Iteration 17500, loss = 6.42806 | |
| I0330 02:17:40.990348 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0545455 | |
| I0330 02:17:40.990370 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.704545 | |
| I0330 02:17:40.990383 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.145455 | |
| I0330 02:17:40.990401 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.36336 (* 0.3 = 1.00901 loss) | |
| I0330 02:17:40.990417 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.10731 (* 0.3 = 0.332194 loss) | |
| I0330 02:17:40.990429 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0909091 | |
| I0330 02:17:40.990442 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0330 02:17:40.990454 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.127273 | |
| I0330 02:17:40.990468 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.4856 (* 0.3 = 1.04568 loss) | |
| I0330 02:17:40.990483 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.1272 (* 0.3 = 0.338161 loss) | |
| I0330 02:17:40.990494 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0909091 | |
| I0330 02:17:40.990507 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.715909 | |
| I0330 02:17:40.990520 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.2 | |
| I0330 02:17:40.990533 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.282 (* 1 = 3.282 loss) | |
| I0330 02:17:40.990547 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.05585 (* 1 = 1.05585 loss) | |
| I0330 02:17:40.990559 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:17:40.990571 10583 solver.cpp:245] Train net output #16: total_confidence = 1.95713e-06 | |
| I0330 02:17:40.990583 10583 sgd_solver.cpp:106] Iteration 17500, lr = 0.01 | |
| I0330 02:19:50.701715 10583 solver.cpp:229] Iteration 18000, loss = 6.40321 | |
| I0330 02:19:50.701863 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145833 | |
| I0330 02:19:50.701884 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 02:19:50.701897 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.25 | |
| I0330 02:19:50.701915 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.11099 (* 0.3 = 0.933296 loss) | |
| I0330 02:19:50.701928 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.935673 (* 0.3 = 0.280702 loss) | |
| I0330 02:19:50.701941 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0625 | |
| I0330 02:19:50.701956 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 02:19:50.701967 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.229167 | |
| I0330 02:19:50.701982 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.15114 (* 0.3 = 0.945342 loss) | |
| I0330 02:19:50.701997 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.947162 (* 0.3 = 0.284149 loss) | |
| I0330 02:19:50.702008 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.125 | |
| I0330 02:19:50.702021 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 02:19:50.702033 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.1875 | |
| I0330 02:19:50.702049 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.95078 (* 1 = 2.95078 loss) | |
| I0330 02:19:50.702064 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.869415 (* 1 = 0.869415 loss) | |
| I0330 02:19:50.702077 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:19:50.702090 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000197928 | |
| I0330 02:19:50.702102 10583 sgd_solver.cpp:106] Iteration 18000, lr = 0.01 | |
| I0330 02:22:00.545068 10583 solver.cpp:229] Iteration 18500, loss = 6.41448 | |
| I0330 02:22:00.545254 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0714286 | |
| I0330 02:22:00.545285 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 02:22:00.545307 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.214286 | |
| I0330 02:22:00.545336 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.11681 (* 0.3 = 0.935044 loss) | |
| I0330 02:22:00.545361 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.812152 (* 0.3 = 0.243646 loss) | |
| I0330 02:22:00.545382 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0714286 | |
| I0330 02:22:00.545397 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 02:22:00.545409 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.214286 | |
| I0330 02:22:00.545424 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.03531 (* 0.3 = 0.910592 loss) | |
| I0330 02:22:00.545439 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.815164 (* 0.3 = 0.244549 loss) | |
| I0330 02:22:00.545451 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.166667 | |
| I0330 02:22:00.545464 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 02:22:00.545476 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.261905 | |
| I0330 02:22:00.545491 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.8432 (* 1 = 2.8432 loss) | |
| I0330 02:22:00.545506 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.769977 (* 1 = 0.769977 loss) | |
| I0330 02:22:00.545518 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:22:00.545531 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000227487 | |
| I0330 02:22:00.545543 10583 sgd_solver.cpp:106] Iteration 18500, lr = 0.01 | |
| I0330 02:24:10.893762 10583 solver.cpp:229] Iteration 19000, loss = 6.3891 | |
| I0330 02:24:10.893923 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.108696 | |
| I0330 02:24:10.893945 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 02:24:10.893959 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.282609 | |
| I0330 02:24:10.893975 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.93446 (* 0.3 = 0.880338 loss) | |
| I0330 02:24:10.893990 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.883257 (* 0.3 = 0.264977 loss) | |
| I0330 02:24:10.894003 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.152174 | |
| I0330 02:24:10.894016 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 02:24:10.894028 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.326087 | |
| I0330 02:24:10.894042 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.8298 (* 0.3 = 0.84894 loss) | |
| I0330 02:24:10.894057 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.867833 (* 0.3 = 0.26035 loss) | |
| I0330 02:24:10.894070 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.108696 | |
| I0330 02:24:10.894083 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0330 02:24:10.894095 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.304348 | |
| I0330 02:24:10.894110 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.82815 (* 1 = 2.82815 loss) | |
| I0330 02:24:10.894125 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.833596 (* 1 = 0.833596 loss) | |
| I0330 02:24:10.894139 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:24:10.894150 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000335385 | |
| I0330 02:24:10.894166 10583 sgd_solver.cpp:106] Iteration 19000, lr = 0.01 | |
| I0330 02:26:21.180776 10583 solver.cpp:229] Iteration 19500, loss = 6.29862 | |
| I0330 02:26:21.180971 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0588235 | |
| I0330 02:26:21.180994 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 02:26:21.181006 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.215686 | |
| I0330 02:26:21.181023 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.07434 (* 0.3 = 0.922303 loss) | |
| I0330 02:26:21.181038 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.928321 (* 0.3 = 0.278496 loss) | |
| I0330 02:26:21.181054 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.137255 | |
| I0330 02:26:21.181072 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 02:26:21.181088 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.254902 | |
| I0330 02:26:21.181103 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.01694 (* 0.3 = 0.905083 loss) | |
| I0330 02:26:21.181118 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.956224 (* 0.3 = 0.286867 loss) | |
| I0330 02:26:21.181130 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0784314 | |
| I0330 02:26:21.181143 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.727273 | |
| I0330 02:26:21.181155 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.235294 | |
| I0330 02:26:21.181174 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.94927 (* 1 = 2.94927 loss) | |
| I0330 02:26:21.181188 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.898064 (* 1 = 0.898064 loss) | |
| I0330 02:26:21.181200 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:26:21.181212 10583 solver.cpp:245] Train net output #16: total_confidence = 1.03552e-05 | |
| I0330 02:26:21.181226 10583 sgd_solver.cpp:106] Iteration 19500, lr = 0.01 | |
| I0330 02:28:31.271493 10583 solver.cpp:338] Iteration 20000, Testing net (#0) | |
| I0330 02:29:01.492352 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 02:29:01.492475 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 02:29:01.492493 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.76 | |
| I0330 02:29:01.492506 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 02:29:01.492523 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:29:01.492539 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:29:01.492552 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 02:29:01.492563 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.76 | |
| I0330 02:29:01.492575 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 02:29:01.492589 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:29:01.492604 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:29:01.492616 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 02:29:01.492629 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.76 | |
| I0330 02:29:01.492640 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 02:29:01.492655 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:29:01.492668 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:29:01.492681 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 02:29:01.492693 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 02:29:01.644359 10583 solver.cpp:229] Iteration 20000, loss = 6.33314 | |
| I0330 02:29:01.644431 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0888889 | |
| I0330 02:29:01.644449 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 02:29:01.644464 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.244444 | |
| I0330 02:29:01.644480 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.15227 (* 0.3 = 0.945682 loss) | |
| I0330 02:29:01.644495 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.890989 (* 0.3 = 0.267297 loss) | |
| I0330 02:29:01.644508 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.133333 | |
| I0330 02:29:01.644521 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 02:29:01.644533 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.311111 | |
| I0330 02:29:01.644548 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.06827 (* 0.3 = 0.920483 loss) | |
| I0330 02:29:01.644562 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.872768 (* 0.3 = 0.26183 loss) | |
| I0330 02:29:01.644574 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.222222 | |
| I0330 02:29:01.644587 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 02:29:01.644599 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.377778 | |
| I0330 02:29:01.644613 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.922 (* 1 = 2.922 loss) | |
| I0330 02:29:01.644628 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.815567 (* 1 = 0.815567 loss) | |
| I0330 02:29:01.644639 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:29:01.644651 10583 solver.cpp:245] Train net output #16: total_confidence = 2.06074e-05 | |
| I0330 02:29:01.644665 10583 sgd_solver.cpp:106] Iteration 20000, lr = 0.01 | |
| I0330 02:31:11.886114 10583 solver.cpp:229] Iteration 20500, loss = 6.31933 | |
| I0330 02:31:11.886284 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0652174 | |
| I0330 02:31:11.886308 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 02:31:11.886323 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.26087 | |
| I0330 02:31:11.886339 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.95514 (* 0.3 = 0.886542 loss) | |
| I0330 02:31:11.886355 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.881492 (* 0.3 = 0.264448 loss) | |
| I0330 02:31:11.886368 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0869565 | |
| I0330 02:31:11.886381 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 02:31:11.886392 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.282609 | |
| I0330 02:31:11.886407 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.94971 (* 0.3 = 0.884914 loss) | |
| I0330 02:31:11.886421 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.908146 (* 0.3 = 0.272444 loss) | |
| I0330 02:31:11.886433 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.108696 | |
| I0330 02:31:11.886446 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 02:31:11.886458 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.326087 | |
| I0330 02:31:11.886473 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.85237 (* 1 = 2.85237 loss) | |
| I0330 02:31:11.886487 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.826255 (* 1 = 0.826255 loss) | |
| I0330 02:31:11.886499 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:31:11.886512 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000109719 | |
| I0330 02:31:11.886524 10583 sgd_solver.cpp:106] Iteration 20500, lr = 0.01 | |
| I0330 02:33:21.705014 10583 solver.cpp:229] Iteration 21000, loss = 6.29526 | |
| I0330 02:33:21.705157 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0330 02:33:21.705176 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 02:33:21.705190 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.2 | |
| I0330 02:33:21.705206 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.20813 (* 0.3 = 0.96244 loss) | |
| I0330 02:33:21.705221 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.938352 (* 0.3 = 0.281506 loss) | |
| I0330 02:33:21.705235 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0888889 | |
| I0330 02:33:21.705247 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0330 02:33:21.705260 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.288889 | |
| I0330 02:33:21.705273 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.07531 (* 0.3 = 0.922592 loss) | |
| I0330 02:33:21.705287 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.845511 (* 0.3 = 0.253653 loss) | |
| I0330 02:33:21.705301 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0888889 | |
| I0330 02:33:21.705312 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 02:33:21.705324 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.266667 | |
| I0330 02:33:21.705338 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.05761 (* 1 = 3.05761 loss) | |
| I0330 02:33:21.705353 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.8487 (* 1 = 0.8487 loss) | |
| I0330 02:33:21.705365 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:33:21.705377 10583 solver.cpp:245] Train net output #16: total_confidence = 3.13307e-05 | |
| I0330 02:33:21.705390 10583 sgd_solver.cpp:106] Iteration 21000, lr = 0.01 | |
| I0330 02:35:31.393710 10583 solver.cpp:229] Iteration 21500, loss = 6.31004 | |
| I0330 02:35:31.393875 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1875 | |
| I0330 02:35:31.393896 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 02:35:31.393910 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.3125 | |
| I0330 02:35:31.393928 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.84769 (* 0.3 = 0.854308 loss) | |
| I0330 02:35:31.393942 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.880226 (* 0.3 = 0.264068 loss) | |
| I0330 02:35:31.393955 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0833333 | |
| I0330 02:35:31.393968 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 02:35:31.393980 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.354167 | |
| I0330 02:35:31.393996 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.72943 (* 0.3 = 0.81883 loss) | |
| I0330 02:35:31.394011 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.809434 (* 0.3 = 0.24283 loss) | |
| I0330 02:35:31.394023 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.1875 | |
| I0330 02:35:31.394037 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 02:35:31.394048 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.395833 | |
| I0330 02:35:31.394062 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.65246 (* 1 = 2.65246 loss) | |
| I0330 02:35:31.394076 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.80788 (* 1 = 0.80788 loss) | |
| I0330 02:35:31.394089 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:35:31.394101 10583 solver.cpp:245] Train net output #16: total_confidence = 3.00169e-05 | |
| I0330 02:35:31.394114 10583 sgd_solver.cpp:106] Iteration 21500, lr = 0.01 | |
| I0330 02:37:41.011801 10583 solver.cpp:229] Iteration 22000, loss = 6.24657 | |
| I0330 02:37:41.011941 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0869565 | |
| I0330 02:37:41.011962 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 02:37:41.011976 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.282609 | |
| I0330 02:37:41.011992 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.16542 (* 0.3 = 0.949627 loss) | |
| I0330 02:37:41.012007 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.917691 (* 0.3 = 0.275307 loss) | |
| I0330 02:37:41.012020 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0652174 | |
| I0330 02:37:41.012032 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 02:37:41.012044 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.23913 | |
| I0330 02:37:41.012058 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.06557 (* 0.3 = 0.919671 loss) | |
| I0330 02:37:41.012073 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.887087 (* 0.3 = 0.266126 loss) | |
| I0330 02:37:41.012084 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0652174 | |
| I0330 02:37:41.012096 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 02:37:41.012109 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.195652 | |
| I0330 02:37:41.012123 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.94592 (* 1 = 2.94592 loss) | |
| I0330 02:37:41.012137 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.845894 (* 1 = 0.845894 loss) | |
| I0330 02:37:41.012150 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:37:41.012163 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000417959 | |
| I0330 02:37:41.012177 10583 sgd_solver.cpp:106] Iteration 22000, lr = 0.01 | |
| I0330 02:39:51.083767 10583 solver.cpp:229] Iteration 22500, loss = 6.19836 | |
| I0330 02:39:51.083900 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1 | |
| I0330 02:39:51.083921 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 02:39:51.083936 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.25 | |
| I0330 02:39:51.083952 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.21113 (* 0.3 = 0.963338 loss) | |
| I0330 02:39:51.083967 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.857486 (* 0.3 = 0.257246 loss) | |
| I0330 02:39:51.083979 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.125 | |
| I0330 02:39:51.083992 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 02:39:51.084004 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.275 | |
| I0330 02:39:51.084018 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.96193 (* 0.3 = 0.888578 loss) | |
| I0330 02:39:51.084033 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.818479 (* 0.3 = 0.245544 loss) | |
| I0330 02:39:51.084045 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.2 | |
| I0330 02:39:51.084056 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 02:39:51.084069 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.275 | |
| I0330 02:39:51.084082 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.92647 (* 1 = 2.92647 loss) | |
| I0330 02:39:51.084096 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.794395 (* 1 = 0.794395 loss) | |
| I0330 02:39:51.084108 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:39:51.084120 10583 solver.cpp:245] Train net output #16: total_confidence = 6.55988e-05 | |
| I0330 02:39:51.084134 10583 sgd_solver.cpp:106] Iteration 22500, lr = 0.01 | |
| I0330 02:42:00.688395 10583 solver.cpp:229] Iteration 23000, loss = 6.14507 | |
| I0330 02:42:00.688606 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.183673 | |
| I0330 02:42:00.688630 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 02:42:00.688644 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.367347 | |
| I0330 02:42:00.688660 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.99449 (* 0.3 = 0.898346 loss) | |
| I0330 02:42:00.688675 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.907056 (* 0.3 = 0.272117 loss) | |
| I0330 02:42:00.688688 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.102041 | |
| I0330 02:42:00.688701 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 02:42:00.688714 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.326531 | |
| I0330 02:42:00.688727 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.958 (* 0.3 = 0.8874 loss) | |
| I0330 02:42:00.688742 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.911104 (* 0.3 = 0.273331 loss) | |
| I0330 02:42:00.688755 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.244898 | |
| I0330 02:42:00.688767 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 02:42:00.688779 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.387755 | |
| I0330 02:42:00.688793 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.85314 (* 1 = 2.85314 loss) | |
| I0330 02:42:00.688807 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.834575 (* 1 = 0.834575 loss) | |
| I0330 02:42:00.688819 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:42:00.688832 10583 solver.cpp:245] Train net output #16: total_confidence = 7.35942e-05 | |
| I0330 02:42:00.688844 10583 sgd_solver.cpp:106] Iteration 23000, lr = 0.01 | |
| I0330 02:44:10.531247 10583 solver.cpp:229] Iteration 23500, loss = 6.12069 | |
| I0330 02:44:10.531402 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.136364 | |
| I0330 02:44:10.531422 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 02:44:10.531437 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.318182 | |
| I0330 02:44:10.531453 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03078 (* 0.3 = 0.909233 loss) | |
| I0330 02:44:10.531468 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.855869 (* 0.3 = 0.256761 loss) | |
| I0330 02:44:10.531481 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.181818 | |
| I0330 02:44:10.531493 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 02:44:10.531505 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.386364 | |
| I0330 02:44:10.531520 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.94379 (* 0.3 = 0.883137 loss) | |
| I0330 02:44:10.531534 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.857304 (* 0.3 = 0.257191 loss) | |
| I0330 02:44:10.531546 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.227273 | |
| I0330 02:44:10.531558 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 02:44:10.531571 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.386364 | |
| I0330 02:44:10.531585 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.01416 (* 1 = 3.01416 loss) | |
| I0330 02:44:10.531599 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.883626 (* 1 = 0.883626 loss) | |
| I0330 02:44:10.531611 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:44:10.531623 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000711273 | |
| I0330 02:44:10.531636 10583 sgd_solver.cpp:106] Iteration 23500, lr = 0.01 | |
| I0330 02:46:20.413043 10583 solver.cpp:229] Iteration 24000, loss = 6.0983 | |
| I0330 02:46:20.413244 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.148936 | |
| I0330 02:46:20.413265 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 02:46:20.413280 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.404255 | |
| I0330 02:46:20.413295 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76522 (* 0.3 = 0.829566 loss) | |
| I0330 02:46:20.413311 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.773505 (* 0.3 = 0.232052 loss) | |
| I0330 02:46:20.413323 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.212766 | |
| I0330 02:46:20.413336 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 02:46:20.413348 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.361702 | |
| I0330 02:46:20.413363 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.61618 (* 0.3 = 0.784853 loss) | |
| I0330 02:46:20.413378 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.739037 (* 0.3 = 0.221711 loss) | |
| I0330 02:46:20.413389 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.191489 | |
| I0330 02:46:20.413401 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 02:46:20.413414 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.510638 | |
| I0330 02:46:20.413429 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.38571 (* 1 = 2.38571 loss) | |
| I0330 02:46:20.413442 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.673515 (* 1 = 0.673515 loss) | |
| I0330 02:46:20.413455 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:46:20.413466 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00212437 | |
| I0330 02:46:20.413480 10583 sgd_solver.cpp:106] Iteration 24000, lr = 0.01 | |
| I0330 02:48:29.672468 10583 solver.cpp:229] Iteration 24500, loss = 6.09116 | |
| I0330 02:48:29.672626 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0330 02:48:29.672654 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 02:48:29.672677 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.371429 | |
| I0330 02:48:29.672703 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.66079 (* 0.3 = 0.798238 loss) | |
| I0330 02:48:29.672729 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.727684 (* 0.3 = 0.218305 loss) | |
| I0330 02:48:29.672750 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.171429 | |
| I0330 02:48:29.672773 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 02:48:29.672796 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.428571 | |
| I0330 02:48:29.672821 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.61866 (* 0.3 = 0.785598 loss) | |
| I0330 02:48:29.672847 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.668818 (* 0.3 = 0.200645 loss) | |
| I0330 02:48:29.672868 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.342857 | |
| I0330 02:48:29.672889 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 02:48:29.672909 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.457143 | |
| I0330 02:48:29.672935 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.32122 (* 1 = 2.32122 loss) | |
| I0330 02:48:29.672960 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.645049 (* 1 = 0.645049 loss) | |
| I0330 02:48:29.672983 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:48:29.673005 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000415813 | |
| I0330 02:48:29.673028 10583 sgd_solver.cpp:106] Iteration 24500, lr = 0.01 | |
| I0330 02:50:38.627158 10583 solver.cpp:338] Iteration 25000, Testing net (#0) | |
| I0330 02:51:08.336804 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 02:51:08.336854 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 02:51:08.336870 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759501 | |
| I0330 02:51:08.336884 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 02:51:08.336901 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:51:08.336916 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:51:08.336928 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 02:51:08.336941 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759501 | |
| I0330 02:51:08.336953 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 02:51:08.336967 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:51:08.336982 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 02:51:08.336993 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 02:51:08.337005 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759501 | |
| I0330 02:51:08.337018 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 02:51:08.337031 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:51:08.337047 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 02:51:08.337059 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 02:51:08.337071 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 02:51:08.487347 10583 solver.cpp:229] Iteration 25000, loss = 6.09095 | |
| I0330 02:51:08.487385 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.177778 | |
| I0330 02:51:08.487401 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 02:51:08.487414 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.266667 | |
| I0330 02:51:08.487429 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03835 (* 0.3 = 0.911506 loss) | |
| I0330 02:51:08.487444 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.94274 (* 0.3 = 0.282822 loss) | |
| I0330 02:51:08.487457 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.222222 | |
| I0330 02:51:08.487469 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 02:51:08.487481 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.355556 | |
| I0330 02:51:08.487496 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.96349 (* 0.3 = 0.889048 loss) | |
| I0330 02:51:08.487510 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.917159 (* 0.3 = 0.275148 loss) | |
| I0330 02:51:08.487522 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.244444 | |
| I0330 02:51:08.487535 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 02:51:08.487547 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.444444 | |
| I0330 02:51:08.487561 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.70947 (* 1 = 2.70947 loss) | |
| I0330 02:51:08.487576 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.789532 (* 1 = 0.789532 loss) | |
| I0330 02:51:08.487587 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:51:08.487599 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000137148 | |
| I0330 02:51:08.487612 10583 sgd_solver.cpp:106] Iteration 25000, lr = 0.01 | |
| I0330 02:53:17.428442 10583 solver.cpp:229] Iteration 25500, loss = 6.00044 | |
| I0330 02:53:17.428592 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.125 | |
| I0330 02:53:17.428613 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 02:53:17.428627 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.225 | |
| I0330 02:53:17.428643 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09329 (* 0.3 = 0.927988 loss) | |
| I0330 02:53:17.428658 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.867613 (* 0.3 = 0.260284 loss) | |
| I0330 02:53:17.428673 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.075 | |
| I0330 02:53:17.428685 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 02:53:17.428697 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.25 | |
| I0330 02:53:17.428711 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.22156 (* 0.3 = 0.966469 loss) | |
| I0330 02:53:17.428725 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.88571 (* 0.3 = 0.265713 loss) | |
| I0330 02:53:17.428737 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.175 | |
| I0330 02:53:17.428750 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 02:53:17.428761 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.3 | |
| I0330 02:53:17.428776 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.06028 (* 1 = 3.06028 loss) | |
| I0330 02:53:17.428791 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.845989 (* 1 = 0.845989 loss) | |
| I0330 02:53:17.428803 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:53:17.428815 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000672717 | |
| I0330 02:53:17.428828 10583 sgd_solver.cpp:106] Iteration 25500, lr = 0.01 | |
| I0330 02:55:26.404912 10583 solver.cpp:229] Iteration 26000, loss = 5.97529 | |
| I0330 02:55:26.405035 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0943396 | |
| I0330 02:55:26.405055 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 02:55:26.405067 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.264151 | |
| I0330 02:55:26.405083 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.17191 (* 0.3 = 0.951572 loss) | |
| I0330 02:55:26.405098 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.981684 (* 0.3 = 0.294505 loss) | |
| I0330 02:55:26.405110 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0943396 | |
| I0330 02:55:26.405122 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0330 02:55:26.405134 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.320755 | |
| I0330 02:55:26.405148 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.04137 (* 0.3 = 0.91241 loss) | |
| I0330 02:55:26.405165 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.957334 (* 0.3 = 0.2872 loss) | |
| I0330 02:55:26.405179 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.264151 | |
| I0330 02:55:26.405190 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 02:55:26.405202 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.45283 | |
| I0330 02:55:26.405217 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.77826 (* 1 = 2.77826 loss) | |
| I0330 02:55:26.405231 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.852798 (* 1 = 0.852798 loss) | |
| I0330 02:55:26.405243 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:55:26.405256 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000471583 | |
| I0330 02:55:26.405267 10583 sgd_solver.cpp:106] Iteration 26000, lr = 0.01 | |
| I0330 02:57:35.588066 10583 solver.cpp:229] Iteration 26500, loss = 5.95172 | |
| I0330 02:57:35.588246 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.162791 | |
| I0330 02:57:35.588276 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 02:57:35.588300 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.27907 | |
| I0330 02:57:35.588330 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.05687 (* 0.3 = 0.917062 loss) | |
| I0330 02:57:35.588353 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.853996 (* 0.3 = 0.256199 loss) | |
| I0330 02:57:35.588367 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.139535 | |
| I0330 02:57:35.588381 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 02:57:35.588392 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.255814 | |
| I0330 02:57:35.588407 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98759 (* 0.3 = 0.896278 loss) | |
| I0330 02:57:35.588421 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.856507 (* 0.3 = 0.256952 loss) | |
| I0330 02:57:35.588433 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.139535 | |
| I0330 02:57:35.588446 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 02:57:35.588459 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.325581 | |
| I0330 02:57:35.588472 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.76676 (* 1 = 2.76676 loss) | |
| I0330 02:57:35.588486 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.779836 (* 1 = 0.779836 loss) | |
| I0330 02:57:35.588500 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:57:35.588511 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000381611 | |
| I0330 02:57:35.588524 10583 sgd_solver.cpp:106] Iteration 26500, lr = 0.01 | |
| I0330 02:59:44.922421 10583 solver.cpp:229] Iteration 27000, loss = 6.00108 | |
| I0330 02:59:44.922582 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1 | |
| I0330 02:59:44.922602 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 02:59:44.922617 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.275 | |
| I0330 02:59:44.922633 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.81498 (* 0.3 = 0.844493 loss) | |
| I0330 02:59:44.922649 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.764412 (* 0.3 = 0.229324 loss) | |
| I0330 02:59:44.922662 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.225 | |
| I0330 02:59:44.922674 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 02:59:44.922686 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.35 | |
| I0330 02:59:44.922701 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.67625 (* 0.3 = 0.802875 loss) | |
| I0330 02:59:44.922715 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.797846 (* 0.3 = 0.239354 loss) | |
| I0330 02:59:44.922729 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.275 | |
| I0330 02:59:44.922740 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0330 02:59:44.922752 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.35 | |
| I0330 02:59:44.922767 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.44522 (* 1 = 2.44522 loss) | |
| I0330 02:59:44.922781 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.695068 (* 1 = 0.695068 loss) | |
| I0330 02:59:44.922794 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 02:59:44.922806 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000567376 | |
| I0330 02:59:44.922819 10583 sgd_solver.cpp:106] Iteration 27000, lr = 0.01 | |
| I0330 03:01:55.343412 10583 solver.cpp:229] Iteration 27500, loss = 5.92132 | |
| I0330 03:01:55.343636 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.152174 | |
| I0330 03:01:55.343658 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 03:01:55.343672 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.326087 | |
| I0330 03:01:55.343689 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09797 (* 0.3 = 0.929392 loss) | |
| I0330 03:01:55.343704 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.947563 (* 0.3 = 0.284269 loss) | |
| I0330 03:01:55.343718 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.195652 | |
| I0330 03:01:55.343729 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 03:01:55.343741 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.347826 | |
| I0330 03:01:55.343756 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.96785 (* 0.3 = 0.890355 loss) | |
| I0330 03:01:55.343770 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.938369 (* 0.3 = 0.281511 loss) | |
| I0330 03:01:55.343783 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.130435 | |
| I0330 03:01:55.343796 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 03:01:55.343807 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.282609 | |
| I0330 03:01:55.343822 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9233 (* 1 = 2.9233 loss) | |
| I0330 03:01:55.343837 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.906836 (* 1 = 0.906836 loss) | |
| I0330 03:01:55.343848 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:01:55.343860 10583 solver.cpp:245] Train net output #16: total_confidence = 4.10529e-05 | |
| I0330 03:01:55.343874 10583 sgd_solver.cpp:106] Iteration 27500, lr = 0.01 | |
| I0330 03:04:05.718001 10583 solver.cpp:229] Iteration 28000, loss = 5.85766 | |
| I0330 03:04:05.718178 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535 | |
| I0330 03:04:05.718199 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 03:04:05.718211 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.325581 | |
| I0330 03:04:05.718228 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.75706 (* 0.3 = 0.827119 loss) | |
| I0330 03:04:05.718243 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.787814 (* 0.3 = 0.236344 loss) | |
| I0330 03:04:05.718256 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.186047 | |
| I0330 03:04:05.718269 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 03:04:05.718281 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.325581 | |
| I0330 03:04:05.718307 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.62164 (* 0.3 = 0.786492 loss) | |
| I0330 03:04:05.718334 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.759699 (* 0.3 = 0.22791 loss) | |
| I0330 03:04:05.718353 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.255814 | |
| I0330 03:04:05.718367 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 03:04:05.718379 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.44186 | |
| I0330 03:04:05.718394 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.38417 (* 1 = 2.38417 loss) | |
| I0330 03:04:05.718408 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.707923 (* 1 = 0.707923 loss) | |
| I0330 03:04:05.718421 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:04:05.718433 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000218632 | |
| I0330 03:04:05.718447 10583 sgd_solver.cpp:106] Iteration 28000, lr = 0.01 | |
| I0330 03:06:15.721453 10583 solver.cpp:229] Iteration 28500, loss = 5.83084 | |
| I0330 03:06:15.721655 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.204545 | |
| I0330 03:06:15.721676 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 03:06:15.721691 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.318182 | |
| I0330 03:06:15.721709 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.87419 (* 0.3 = 0.862257 loss) | |
| I0330 03:06:15.721724 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.837546 (* 0.3 = 0.251264 loss) | |
| I0330 03:06:15.721735 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.159091 | |
| I0330 03:06:15.721748 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 03:06:15.721760 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.295455 | |
| I0330 03:06:15.721774 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85693 (* 0.3 = 0.857079 loss) | |
| I0330 03:06:15.721789 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.813396 (* 0.3 = 0.244019 loss) | |
| I0330 03:06:15.721801 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.295455 | |
| I0330 03:06:15.721814 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 03:06:15.721827 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.522727 | |
| I0330 03:06:15.721840 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.47052 (* 1 = 2.47052 loss) | |
| I0330 03:06:15.721855 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.684147 (* 1 = 0.684147 loss) | |
| I0330 03:06:15.721868 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:06:15.721880 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00132636 | |
| I0330 03:06:15.721894 10583 sgd_solver.cpp:106] Iteration 28500, lr = 0.01 | |
| I0330 03:08:25.505940 10583 solver.cpp:229] Iteration 29000, loss = 5.83597 | |
| I0330 03:08:25.506089 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.125 | |
| I0330 03:08:25.506109 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 03:08:25.506124 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.35 | |
| I0330 03:08:25.506140 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.66042 (* 0.3 = 1.09812 loss) | |
| I0330 03:08:25.506155 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.04767 (* 0.3 = 0.314301 loss) | |
| I0330 03:08:25.506171 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.075 | |
| I0330 03:08:25.506184 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 03:08:25.506197 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.3 | |
| I0330 03:08:25.506212 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.68075 (* 0.3 = 1.10422 loss) | |
| I0330 03:08:25.506227 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.992753 (* 0.3 = 0.297826 loss) | |
| I0330 03:08:25.506239 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.25 | |
| I0330 03:08:25.506252 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 03:08:25.506263 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.5 | |
| I0330 03:08:25.506278 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.17489 (* 1 = 3.17489 loss) | |
| I0330 03:08:25.506292 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.860902 (* 1 = 0.860902 loss) | |
| I0330 03:08:25.506305 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:08:25.506317 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000550781 | |
| I0330 03:08:25.506330 10583 sgd_solver.cpp:106] Iteration 29000, lr = 0.01 | |
| I0330 03:10:35.690215 10583 solver.cpp:229] Iteration 29500, loss = 5.78308 | |
| I0330 03:10:35.690587 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.208333 | |
| I0330 03:10:35.690608 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 03:10:35.690623 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.354167 | |
| I0330 03:10:35.690640 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.02997 (* 0.3 = 0.908991 loss) | |
| I0330 03:10:35.690656 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.887377 (* 0.3 = 0.266213 loss) | |
| I0330 03:10:35.690668 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0330 03:10:35.690681 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 03:10:35.690693 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.416667 | |
| I0330 03:10:35.690707 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.70298 (* 0.3 = 0.810894 loss) | |
| I0330 03:10:35.690722 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.80646 (* 0.3 = 0.241938 loss) | |
| I0330 03:10:35.690734 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.229167 | |
| I0330 03:10:35.690747 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 03:10:35.690759 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.416667 | |
| I0330 03:10:35.690774 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.79142 (* 1 = 2.79142 loss) | |
| I0330 03:10:35.690788 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.82014 (* 1 = 0.82014 loss) | |
| I0330 03:10:35.690801 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:10:35.690812 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00065356 | |
| I0330 03:10:35.690826 10583 sgd_solver.cpp:106] Iteration 29500, lr = 0.01 | |
| I0330 03:12:45.302430 10583 solver.cpp:338] Iteration 30000, Testing net (#0) | |
| I0330 03:13:15.175633 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 03:13:15.175698 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 03:13:15.175714 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760001 | |
| I0330 03:13:15.175727 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 03:13:15.175745 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:13:15.175760 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:13:15.175772 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 03:13:15.175783 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760001 | |
| I0330 03:13:15.175796 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 03:13:15.175809 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:13:15.175823 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:13:15.175835 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 03:13:15.175848 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760001 | |
| I0330 03:13:15.175858 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 03:13:15.175873 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:13:15.175887 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:13:15.175899 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 03:13:15.175911 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 03:13:15.327878 10583 solver.cpp:229] Iteration 30000, loss = 5.69804 | |
| I0330 03:13:15.328032 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.146341 | |
| I0330 03:13:15.328053 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 03:13:15.328066 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.292683 | |
| I0330 03:13:15.328083 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09602 (* 0.3 = 0.928807 loss) | |
| I0330 03:13:15.328096 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.854591 (* 0.3 = 0.256377 loss) | |
| I0330 03:13:15.328109 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.219512 | |
| I0330 03:13:15.328122 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 03:13:15.328135 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.390244 | |
| I0330 03:13:15.328147 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85108 (* 0.3 = 0.855323 loss) | |
| I0330 03:13:15.328164 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.806604 (* 0.3 = 0.241981 loss) | |
| I0330 03:13:15.328179 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.292683 | |
| I0330 03:13:15.328192 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0330 03:13:15.328203 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.487805 | |
| I0330 03:13:15.328218 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.40789 (* 1 = 2.40789 loss) | |
| I0330 03:13:15.328233 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.659126 (* 1 = 0.659126 loss) | |
| I0330 03:13:15.328244 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:13:15.328256 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000688415 | |
| I0330 03:13:15.328269 10583 sgd_solver.cpp:106] Iteration 30000, lr = 0.01 | |
| I0330 03:15:24.505162 10583 solver.cpp:229] Iteration 30500, loss = 5.73712 | |
| I0330 03:15:24.505323 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.128205 | |
| I0330 03:15:24.505343 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 03:15:24.505357 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.179487 | |
| I0330 03:15:24.505374 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.66584 (* 0.3 = 1.09975 loss) | |
| I0330 03:15:24.505389 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.00234 (* 0.3 = 0.300703 loss) | |
| I0330 03:15:24.505403 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.153846 | |
| I0330 03:15:24.505414 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 03:15:24.505426 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.25641 | |
| I0330 03:15:24.505441 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.62887 (* 0.3 = 1.08866 loss) | |
| I0330 03:15:24.505456 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.0201 (* 0.3 = 0.306029 loss) | |
| I0330 03:15:24.505468 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.153846 | |
| I0330 03:15:24.505481 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 03:15:24.505492 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.358974 | |
| I0330 03:15:24.505506 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.09656 (* 1 = 3.09656 loss) | |
| I0330 03:15:24.505522 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.844311 (* 1 = 0.844311 loss) | |
| I0330 03:15:24.505533 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:15:24.505545 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00112581 | |
| I0330 03:15:24.505558 10583 sgd_solver.cpp:106] Iteration 30500, lr = 0.01 | |
| I0330 03:17:33.802902 10583 solver.cpp:229] Iteration 31000, loss = 5.71364 | |
| I0330 03:17:33.803053 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.169811 | |
| I0330 03:17:33.803074 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0330 03:17:33.803088 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.415094 | |
| I0330 03:17:33.803104 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76972 (* 0.3 = 0.830917 loss) | |
| I0330 03:17:33.803119 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.994235 (* 0.3 = 0.29827 loss) | |
| I0330 03:17:33.803131 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.207547 | |
| I0330 03:17:33.803144 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0330 03:17:33.803156 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.433962 | |
| I0330 03:17:33.803174 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.75962 (* 0.3 = 0.827885 loss) | |
| I0330 03:17:33.803189 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.953121 (* 0.3 = 0.285936 loss) | |
| I0330 03:17:33.803200 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.283019 | |
| I0330 03:17:33.803213 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 03:17:33.803225 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.45283 | |
| I0330 03:17:33.803239 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.54575 (* 1 = 2.54575 loss) | |
| I0330 03:17:33.803254 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.870677 (* 1 = 0.870677 loss) | |
| I0330 03:17:33.803267 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:17:33.803278 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000809857 | |
| I0330 03:17:33.803292 10583 sgd_solver.cpp:106] Iteration 31000, lr = 0.01 | |
| I0330 03:19:43.126396 10583 solver.cpp:229] Iteration 31500, loss = 5.62782 | |
| I0330 03:19:43.126546 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.108696 | |
| I0330 03:19:43.126567 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 03:19:43.126581 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.282609 | |
| I0330 03:19:43.126598 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.4528 (* 0.3 = 1.03584 loss) | |
| I0330 03:19:43.126613 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.0244 (* 0.3 = 0.307321 loss) | |
| I0330 03:19:43.126626 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.130435 | |
| I0330 03:19:43.126638 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 03:19:43.126651 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.304348 | |
| I0330 03:19:43.126664 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.23546 (* 0.3 = 0.970638 loss) | |
| I0330 03:19:43.126678 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.935604 (* 0.3 = 0.280681 loss) | |
| I0330 03:19:43.126693 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.195652 | |
| I0330 03:19:43.126706 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 03:19:43.126718 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.434783 | |
| I0330 03:19:43.126734 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.04713 (* 1 = 3.04713 loss) | |
| I0330 03:19:43.126747 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.859275 (* 1 = 0.859275 loss) | |
| I0330 03:19:43.126760 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:19:43.126771 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00285272 | |
| I0330 03:19:43.126785 10583 sgd_solver.cpp:106] Iteration 31500, lr = 0.01 | |
| I0330 03:21:52.432822 10583 solver.cpp:229] Iteration 32000, loss = 5.6207 | |
| I0330 03:21:52.432961 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.283019 | |
| I0330 03:21:52.432984 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 03:21:52.432998 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509434 | |
| I0330 03:21:52.433015 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.56695 (* 0.3 = 0.770085 loss) | |
| I0330 03:21:52.433030 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.798394 (* 0.3 = 0.239518 loss) | |
| I0330 03:21:52.433043 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.358491 | |
| I0330 03:21:52.433055 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 03:21:52.433068 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.603774 | |
| I0330 03:21:52.433081 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.31326 (* 0.3 = 0.693977 loss) | |
| I0330 03:21:52.433095 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.725336 (* 0.3 = 0.217601 loss) | |
| I0330 03:21:52.433109 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.339623 | |
| I0330 03:21:52.433120 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 03:21:52.433131 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.716981 | |
| I0330 03:21:52.433146 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.98152 (* 1 = 1.98152 loss) | |
| I0330 03:21:52.433161 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.606382 (* 1 = 0.606382 loss) | |
| I0330 03:21:52.433172 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:21:52.433184 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00125908 | |
| I0330 03:21:52.433197 10583 sgd_solver.cpp:106] Iteration 32000, lr = 0.01 | |
| I0330 03:24:01.834766 10583 solver.cpp:229] Iteration 32500, loss = 5.55867 | |
| I0330 03:24:01.834923 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1 | |
| I0330 03:24:01.834944 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 03:24:01.834957 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.275 | |
| I0330 03:24:01.834975 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.13214 (* 0.3 = 0.939644 loss) | |
| I0330 03:24:01.834990 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.928757 (* 0.3 = 0.278627 loss) | |
| I0330 03:24:01.835003 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.175 | |
| I0330 03:24:01.835016 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 03:24:01.835027 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0330 03:24:01.835053 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.87695 (* 0.3 = 0.863086 loss) | |
| I0330 03:24:01.835069 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.903572 (* 0.3 = 0.271072 loss) | |
| I0330 03:24:01.835083 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.2 | |
| I0330 03:24:01.835095 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 03:24:01.835108 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.425 | |
| I0330 03:24:01.835121 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.7146 (* 1 = 2.7146 loss) | |
| I0330 03:24:01.835135 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.819137 (* 1 = 0.819137 loss) | |
| I0330 03:24:01.835149 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:24:01.835162 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00131826 | |
| I0330 03:24:01.835176 10583 sgd_solver.cpp:106] Iteration 32500, lr = 0.01 | |
| I0330 03:26:11.123893 10583 solver.cpp:229] Iteration 33000, loss = 5.54857 | |
| I0330 03:26:11.124086 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0330 03:26:11.124107 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 03:26:11.124120 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.44 | |
| I0330 03:26:11.124138 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76387 (* 0.3 = 0.82916 loss) | |
| I0330 03:26:11.124153 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.853513 (* 0.3 = 0.256054 loss) | |
| I0330 03:26:11.124169 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22 | |
| I0330 03:26:11.124182 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 03:26:11.124194 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0330 03:26:11.124209 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.83896 (* 0.3 = 0.851689 loss) | |
| I0330 03:26:11.124224 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.864275 (* 0.3 = 0.259283 loss) | |
| I0330 03:26:11.124236 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.24 | |
| I0330 03:26:11.124249 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 03:26:11.124261 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.52 | |
| I0330 03:26:11.124276 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.34531 (* 1 = 2.34531 loss) | |
| I0330 03:26:11.124291 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.739927 (* 1 = 0.739927 loss) | |
| I0330 03:26:11.124303 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:26:11.124316 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000523572 | |
| I0330 03:26:11.124330 10583 sgd_solver.cpp:106] Iteration 33000, lr = 0.01 | |
| I0330 03:28:20.465715 10583 solver.cpp:229] Iteration 33500, loss = 5.57474 | |
| I0330 03:28:20.465869 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1 | |
| I0330 03:28:20.465890 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 03:28:20.465904 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.2 | |
| I0330 03:28:20.465920 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.46719 (* 0.3 = 1.04016 loss) | |
| I0330 03:28:20.465935 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.06042 (* 0.3 = 0.318126 loss) | |
| I0330 03:28:20.465948 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.08 | |
| I0330 03:28:20.465960 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0330 03:28:20.465973 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.12 | |
| I0330 03:28:20.465987 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.13092 (* 0.3 = 0.939277 loss) | |
| I0330 03:28:20.466002 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.985978 (* 0.3 = 0.295793 loss) | |
| I0330 03:28:20.466014 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.16 | |
| I0330 03:28:20.466027 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0330 03:28:20.466039 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.26 | |
| I0330 03:28:20.466053 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.0481 (* 1 = 3.0481 loss) | |
| I0330 03:28:20.466068 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.910363 (* 1 = 0.910363 loss) | |
| I0330 03:28:20.466080 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:28:20.466092 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00100237 | |
| I0330 03:28:20.466105 10583 sgd_solver.cpp:106] Iteration 33500, lr = 0.01 | |
| I0330 03:30:29.827690 10583 solver.cpp:229] Iteration 34000, loss = 5.53289 | |
| I0330 03:30:29.827836 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.142857 | |
| I0330 03:30:29.827857 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 03:30:29.827870 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.339286 | |
| I0330 03:30:29.827886 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.2496 (* 0.3 = 0.974879 loss) | |
| I0330 03:30:29.827901 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.04687 (* 0.3 = 0.31406 loss) | |
| I0330 03:30:29.827914 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.196429 | |
| I0330 03:30:29.827927 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 03:30:29.827939 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.321429 | |
| I0330 03:30:29.827953 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.0436 (* 0.3 = 0.913081 loss) | |
| I0330 03:30:29.827968 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.977344 (* 0.3 = 0.293203 loss) | |
| I0330 03:30:29.827980 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.214286 | |
| I0330 03:30:29.827993 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0330 03:30:29.828006 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.428571 | |
| I0330 03:30:29.828019 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.82705 (* 1 = 2.82705 loss) | |
| I0330 03:30:29.828033 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.905583 (* 1 = 0.905583 loss) | |
| I0330 03:30:29.828047 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:30:29.828058 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0020392 | |
| I0330 03:30:29.828070 10583 sgd_solver.cpp:106] Iteration 34000, lr = 0.01 | |
| I0330 03:32:39.212354 10583 solver.cpp:229] Iteration 34500, loss = 5.46768 | |
| I0330 03:32:39.212484 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.219512 | |
| I0330 03:32:39.212504 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 03:32:39.212517 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.487805 | |
| I0330 03:32:39.212534 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41974 (* 0.3 = 0.725922 loss) | |
| I0330 03:32:39.212550 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.694204 (* 0.3 = 0.208261 loss) | |
| I0330 03:32:39.212563 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.341463 | |
| I0330 03:32:39.212575 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 03:32:39.212587 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.609756 | |
| I0330 03:32:39.212601 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.2278 (* 0.3 = 0.668339 loss) | |
| I0330 03:32:39.212615 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.612562 (* 0.3 = 0.183769 loss) | |
| I0330 03:32:39.212628 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.365854 | |
| I0330 03:32:39.212641 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 03:32:39.212652 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.658537 | |
| I0330 03:32:39.212667 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87954 (* 1 = 1.87954 loss) | |
| I0330 03:32:39.212682 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.565961 (* 1 = 0.565961 loss) | |
| I0330 03:32:39.212693 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:32:39.212705 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00269729 | |
| I0330 03:32:39.212718 10583 sgd_solver.cpp:106] Iteration 34500, lr = 0.01 | |
| I0330 03:34:48.571259 10583 solver.cpp:338] Iteration 35000, Testing net (#0) | |
| I0330 03:35:18.440255 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 03:35:18.440307 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 03:35:18.440323 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759273 | |
| I0330 03:35:18.440336 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 03:35:18.440353 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:35:18.440369 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:35:18.440382 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 03:35:18.440393 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759273 | |
| I0330 03:35:18.440405 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 03:35:18.440419 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:35:18.440434 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:35:18.440446 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 03:35:18.440459 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759273 | |
| I0330 03:35:18.440470 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 03:35:18.440485 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:35:18.440498 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:35:18.440511 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 03:35:18.440522 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 03:35:18.591728 10583 solver.cpp:229] Iteration 35000, loss = 5.41747 | |
| I0330 03:35:18.591853 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195122 | |
| I0330 03:35:18.591872 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 03:35:18.591886 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.414634 | |
| I0330 03:35:18.591902 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76504 (* 0.3 = 0.829511 loss) | |
| I0330 03:35:18.591917 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.821079 (* 0.3 = 0.246324 loss) | |
| I0330 03:35:18.591930 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.268293 | |
| I0330 03:35:18.591943 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 03:35:18.591954 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.439024 | |
| I0330 03:35:18.591969 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.61977 (* 0.3 = 0.78593 loss) | |
| I0330 03:35:18.591982 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.831059 (* 0.3 = 0.249318 loss) | |
| I0330 03:35:18.591995 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.341463 | |
| I0330 03:35:18.592007 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0330 03:35:18.592020 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.487805 | |
| I0330 03:35:18.592033 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.35939 (* 1 = 2.35939 loss) | |
| I0330 03:35:18.592048 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.764576 (* 1 = 0.764576 loss) | |
| I0330 03:35:18.592061 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:35:18.592072 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00467775 | |
| I0330 03:35:18.592084 10583 sgd_solver.cpp:106] Iteration 35000, lr = 0.01 | |
| I0330 03:37:27.937166 10583 solver.cpp:229] Iteration 35500, loss = 5.39315 | |
| I0330 03:37:27.937368 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0697674 | |
| I0330 03:37:27.937391 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 03:37:27.937403 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.372093 | |
| I0330 03:37:27.937420 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.40338 (* 0.3 = 1.02102 loss) | |
| I0330 03:37:27.937435 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.948999 (* 0.3 = 0.2847 loss) | |
| I0330 03:37:27.937448 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.139535 | |
| I0330 03:37:27.937461 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 03:37:27.937474 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.372093 | |
| I0330 03:37:27.937487 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.22896 (* 0.3 = 0.968687 loss) | |
| I0330 03:37:27.937501 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.890704 (* 0.3 = 0.267211 loss) | |
| I0330 03:37:27.937515 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.232558 | |
| I0330 03:37:27.937528 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0330 03:37:27.937541 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.488372 | |
| I0330 03:37:27.937554 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9749 (* 1 = 2.9749 loss) | |
| I0330 03:37:27.937568 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.776298 (* 1 = 0.776298 loss) | |
| I0330 03:37:27.937582 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:37:27.937593 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0042267 | |
| I0330 03:37:27.937607 10583 sgd_solver.cpp:106] Iteration 35500, lr = 0.01 | |
| I0330 03:39:37.309311 10583 solver.cpp:229] Iteration 36000, loss = 5.37236 | |
| I0330 03:39:37.309427 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.115385 | |
| I0330 03:39:37.309447 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 03:39:37.309459 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.269231 | |
| I0330 03:39:37.309476 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.13962 (* 0.3 = 0.941887 loss) | |
| I0330 03:39:37.309490 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.985268 (* 0.3 = 0.29558 loss) | |
| I0330 03:39:37.309504 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.153846 | |
| I0330 03:39:37.309515 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 03:39:37.309527 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.384615 | |
| I0330 03:39:37.309543 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.05752 (* 0.3 = 0.917255 loss) | |
| I0330 03:39:37.309557 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.947563 (* 0.3 = 0.284269 loss) | |
| I0330 03:39:37.309569 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.173077 | |
| I0330 03:39:37.309581 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0330 03:39:37.309593 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.480769 | |
| I0330 03:39:37.309607 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.77185 (* 1 = 2.77185 loss) | |
| I0330 03:39:37.309622 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.853121 (* 1 = 0.853121 loss) | |
| I0330 03:39:37.309634 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:39:37.309645 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00664612 | |
| I0330 03:39:37.309659 10583 sgd_solver.cpp:106] Iteration 36000, lr = 0.01 | |
| I0330 03:41:46.683632 10583 solver.cpp:229] Iteration 36500, loss = 5.29577 | |
| I0330 03:41:46.683784 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0714286 | |
| I0330 03:41:46.683805 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 03:41:46.683818 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.357143 | |
| I0330 03:41:46.683835 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.05404 (* 0.3 = 0.916212 loss) | |
| I0330 03:41:46.683850 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.821887 (* 0.3 = 0.246566 loss) | |
| I0330 03:41:46.683861 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.119048 | |
| I0330 03:41:46.683874 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 03:41:46.683887 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.357143 | |
| I0330 03:41:46.683900 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.97362 (* 0.3 = 0.892087 loss) | |
| I0330 03:41:46.683915 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.806403 (* 0.3 = 0.241921 loss) | |
| I0330 03:41:46.683928 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.238095 | |
| I0330 03:41:46.683939 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0330 03:41:46.683951 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.5 | |
| I0330 03:41:46.683965 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.54529 (* 1 = 2.54529 loss) | |
| I0330 03:41:46.683979 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.711412 (* 1 = 0.711412 loss) | |
| I0330 03:41:46.683992 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:41:46.684003 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000634799 | |
| I0330 03:41:46.684016 10583 sgd_solver.cpp:106] Iteration 36500, lr = 0.01 | |
| I0330 03:43:56.223076 10583 solver.cpp:229] Iteration 37000, loss = 5.31702 | |
| I0330 03:43:56.223232 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.136364 | |
| I0330 03:43:56.223255 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 03:43:56.223268 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0330 03:43:56.223285 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03374 (* 0.3 = 0.910122 loss) | |
| I0330 03:43:56.223300 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.892563 (* 0.3 = 0.267769 loss) | |
| I0330 03:43:56.223314 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0909091 | |
| I0330 03:43:56.223326 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 03:43:56.223340 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.363636 | |
| I0330 03:43:56.223352 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.00082 (* 0.3 = 0.900247 loss) | |
| I0330 03:43:56.223367 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.864712 (* 0.3 = 0.259414 loss) | |
| I0330 03:43:56.223379 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.181818 | |
| I0330 03:43:56.223392 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 03:43:56.223403 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.522727 | |
| I0330 03:43:56.223417 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.54059 (* 1 = 2.54059 loss) | |
| I0330 03:43:56.223431 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.749432 (* 1 = 0.749432 loss) | |
| I0330 03:43:56.223443 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:43:56.223455 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000278069 | |
| I0330 03:43:56.223469 10583 sgd_solver.cpp:106] Iteration 37000, lr = 0.01 | |
| I0330 03:46:05.842399 10583 solver.cpp:229] Iteration 37500, loss = 5.2626 | |
| I0330 03:46:05.842552 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1 | |
| I0330 03:46:05.842573 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 03:46:05.842586 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.34 | |
| I0330 03:46:05.842602 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.88096 (* 0.3 = 0.864289 loss) | |
| I0330 03:46:05.842617 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.896508 (* 0.3 = 0.268952 loss) | |
| I0330 03:46:05.842629 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22 | |
| I0330 03:46:05.842641 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 03:46:05.842653 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.38 | |
| I0330 03:46:05.842667 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.75995 (* 0.3 = 0.827985 loss) | |
| I0330 03:46:05.842681 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.893822 (* 0.3 = 0.268147 loss) | |
| I0330 03:46:05.842694 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.3 | |
| I0330 03:46:05.842705 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 03:46:05.842717 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.5 | |
| I0330 03:46:05.842731 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.36289 (* 1 = 2.36289 loss) | |
| I0330 03:46:05.842746 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.733642 (* 1 = 0.733642 loss) | |
| I0330 03:46:05.842757 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:46:05.842769 10583 solver.cpp:245] Train net output #16: total_confidence = 0.000647516 | |
| I0330 03:46:05.842782 10583 sgd_solver.cpp:106] Iteration 37500, lr = 0.01 | |
| I0330 03:48:15.235052 10583 solver.cpp:229] Iteration 38000, loss = 5.25165 | |
| I0330 03:48:15.235172 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0330 03:48:15.235191 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 03:48:15.235205 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.309091 | |
| I0330 03:48:15.235229 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.90683 (* 0.3 = 0.872049 loss) | |
| I0330 03:48:15.235249 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.944636 (* 0.3 = 0.283391 loss) | |
| I0330 03:48:15.235262 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.218182 | |
| I0330 03:48:15.235275 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 03:48:15.235287 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.418182 | |
| I0330 03:48:15.235301 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85716 (* 0.3 = 0.857149 loss) | |
| I0330 03:48:15.235316 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.931382 (* 0.3 = 0.279415 loss) | |
| I0330 03:48:15.235329 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.236364 | |
| I0330 03:48:15.235342 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0330 03:48:15.235353 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.509091 | |
| I0330 03:48:15.235368 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.50018 (* 1 = 2.50018 loss) | |
| I0330 03:48:15.235383 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.804913 (* 1 = 0.804913 loss) | |
| I0330 03:48:15.235394 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:48:15.235406 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00879229 | |
| I0330 03:48:15.235420 10583 sgd_solver.cpp:106] Iteration 38000, lr = 0.01 | |
| I0330 03:50:24.894850 10583 solver.cpp:229] Iteration 38500, loss = 5.25207 | |
| I0330 03:50:24.895027 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.119048 | |
| I0330 03:50:24.895048 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 03:50:24.895062 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.380952 | |
| I0330 03:50:24.895078 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.91639 (* 0.3 = 0.874917 loss) | |
| I0330 03:50:24.895093 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.781135 (* 0.3 = 0.23434 loss) | |
| I0330 03:50:24.895107 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.214286 | |
| I0330 03:50:24.895118 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 03:50:24.895130 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.285714 | |
| I0330 03:50:24.895144 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.95491 (* 0.3 = 0.886472 loss) | |
| I0330 03:50:24.895160 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.872696 (* 0.3 = 0.261809 loss) | |
| I0330 03:50:24.895174 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.285714 | |
| I0330 03:50:24.895186 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 03:50:24.895198 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.428571 | |
| I0330 03:50:24.895213 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.64952 (* 1 = 2.64952 loss) | |
| I0330 03:50:24.895227 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.778949 (* 1 = 0.778949 loss) | |
| I0330 03:50:24.895239 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:50:24.895251 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00902433 | |
| I0330 03:50:24.895264 10583 sgd_solver.cpp:106] Iteration 38500, lr = 0.01 | |
| I0330 03:52:34.327033 10583 solver.cpp:229] Iteration 39000, loss = 5.18229 | |
| I0330 03:52:34.327191 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0330 03:52:34.327213 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 03:52:34.327225 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0330 03:52:34.327242 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68936 (* 0.3 = 0.806809 loss) | |
| I0330 03:52:34.327257 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.730196 (* 0.3 = 0.219059 loss) | |
| I0330 03:52:34.327270 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.181818 | |
| I0330 03:52:34.327281 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 03:52:34.327294 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.431818 | |
| I0330 03:52:34.327309 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.65003 (* 0.3 = 0.795008 loss) | |
| I0330 03:52:34.327324 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.709617 (* 0.3 = 0.212885 loss) | |
| I0330 03:52:34.327337 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.227273 | |
| I0330 03:52:34.327349 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 03:52:34.327361 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.568182 | |
| I0330 03:52:34.327375 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.29046 (* 1 = 2.29046 loss) | |
| I0330 03:52:34.327390 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.650535 (* 1 = 0.650535 loss) | |
| I0330 03:52:34.327402 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:52:34.327414 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0183528 | |
| I0330 03:52:34.327428 10583 sgd_solver.cpp:106] Iteration 39000, lr = 0.01 | |
| I0330 03:54:43.696732 10583 solver.cpp:229] Iteration 39500, loss = 5.1832 | |
| I0330 03:54:43.696884 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.294118 | |
| I0330 03:54:43.696905 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 03:54:43.696918 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.411765 | |
| I0330 03:54:43.696934 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.06458 (* 0.3 = 0.919375 loss) | |
| I0330 03:54:43.696949 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.911349 (* 0.3 = 0.273405 loss) | |
| I0330 03:54:43.696962 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.254902 | |
| I0330 03:54:43.696975 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 03:54:43.696987 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.431373 | |
| I0330 03:54:43.697001 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.7929 (* 0.3 = 0.83787 loss) | |
| I0330 03:54:43.697016 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.869457 (* 0.3 = 0.260837 loss) | |
| I0330 03:54:43.697028 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.294118 | |
| I0330 03:54:43.697041 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 03:54:43.697052 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.588235 | |
| I0330 03:54:43.697067 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.50334 (* 1 = 2.50334 loss) | |
| I0330 03:54:43.697080 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.759519 (* 1 = 0.759519 loss) | |
| I0330 03:54:43.697093 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:54:43.697104 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00625003 | |
| I0330 03:54:43.697118 10583 sgd_solver.cpp:106] Iteration 39500, lr = 0.01 | |
| I0330 03:56:53.613200 10583 solver.cpp:338] Iteration 40000, Testing net (#0) | |
| I0330 03:57:23.841809 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 03:57:23.841935 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 03:57:23.841954 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760091 | |
| I0330 03:57:23.841967 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 03:57:23.841985 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:57:23.842000 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:57:23.842012 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 03:57:23.842025 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760091 | |
| I0330 03:57:23.842036 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 03:57:23.842052 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:57:23.842067 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 03:57:23.842078 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 03:57:23.842090 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760091 | |
| I0330 03:57:23.842103 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 03:57:23.842116 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:57:23.842130 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 03:57:23.842144 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 03:57:23.842154 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 03:57:23.995301 10583 solver.cpp:229] Iteration 40000, loss = 5.1561 | |
| I0330 03:57:23.995380 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.207547 | |
| I0330 03:57:23.995399 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 03:57:23.995412 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.396226 | |
| I0330 03:57:23.995430 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68184 (* 0.3 = 0.804552 loss) | |
| I0330 03:57:23.995445 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.843838 (* 0.3 = 0.253152 loss) | |
| I0330 03:57:23.995458 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.301887 | |
| I0330 03:57:23.995471 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 03:57:23.995483 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.528302 | |
| I0330 03:57:23.995498 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.54014 (* 0.3 = 0.762043 loss) | |
| I0330 03:57:23.995513 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.807627 (* 0.3 = 0.242288 loss) | |
| I0330 03:57:23.995527 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.320755 | |
| I0330 03:57:23.995538 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 03:57:23.995550 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.54717 | |
| I0330 03:57:23.995564 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.22302 (* 1 = 2.22302 loss) | |
| I0330 03:57:23.995579 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.698972 (* 1 = 0.698972 loss) | |
| I0330 03:57:23.995591 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:57:23.995604 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0150856 | |
| I0330 03:57:23.995617 10583 sgd_solver.cpp:106] Iteration 40000, lr = 0.01 | |
| I0330 03:59:33.419347 10583 solver.cpp:229] Iteration 40500, loss = 5.10681 | |
| I0330 03:59:33.419478 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.227273 | |
| I0330 03:59:33.419497 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 03:59:33.419510 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.409091 | |
| I0330 03:59:33.419526 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.63347 (* 0.3 = 0.79004 loss) | |
| I0330 03:59:33.419541 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.718712 (* 0.3 = 0.215614 loss) | |
| I0330 03:59:33.419554 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.227273 | |
| I0330 03:59:33.419566 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 03:59:33.419579 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.590909 | |
| I0330 03:59:33.419592 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.47571 (* 0.3 = 0.742714 loss) | |
| I0330 03:59:33.419606 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.709658 (* 0.3 = 0.212897 loss) | |
| I0330 03:59:33.419618 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.386364 | |
| I0330 03:59:33.419631 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 03:59:33.419643 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.704545 | |
| I0330 03:59:33.419657 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.88913 (* 1 = 1.88913 loss) | |
| I0330 03:59:33.419670 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.515394 (* 1 = 0.515394 loss) | |
| I0330 03:59:33.419683 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 03:59:33.419694 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0110214 | |
| I0330 03:59:33.419708 10583 sgd_solver.cpp:106] Iteration 40500, lr = 0.01 | |
| I0330 04:01:42.454735 10583 solver.cpp:229] Iteration 41000, loss = 5.01084 | |
| I0330 04:01:42.454859 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.204545 | |
| I0330 04:01:42.454879 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 04:01:42.454891 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0330 04:01:42.454907 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70996 (* 0.3 = 0.812989 loss) | |
| I0330 04:01:42.454922 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.803782 (* 0.3 = 0.241135 loss) | |
| I0330 04:01:42.454934 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.272727 | |
| I0330 04:01:42.454946 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 04:01:42.454958 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.522727 | |
| I0330 04:01:42.454972 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.46768 (* 0.3 = 0.740303 loss) | |
| I0330 04:01:42.454999 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.695691 (* 0.3 = 0.208707 loss) | |
| I0330 04:01:42.455013 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.409091 | |
| I0330 04:01:42.455026 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 04:01:42.455039 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.727273 | |
| I0330 04:01:42.455052 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.90849 (* 1 = 1.90849 loss) | |
| I0330 04:01:42.455067 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.540113 (* 1 = 0.540113 loss) | |
| I0330 04:01:42.455080 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:01:42.455091 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00597293 | |
| I0330 04:01:42.455104 10583 sgd_solver.cpp:106] Iteration 41000, lr = 0.01 | |
| I0330 04:03:51.368062 10583 solver.cpp:229] Iteration 41500, loss = 5.03678 | |
| I0330 04:03:51.368190 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0330 04:03:51.368211 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 04:03:51.368223 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.311111 | |
| I0330 04:03:51.368239 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.33841 (* 0.3 = 1.00152 loss) | |
| I0330 04:03:51.368254 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.93771 (* 0.3 = 0.281313 loss) | |
| I0330 04:03:51.368268 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.266667 | |
| I0330 04:03:51.368279 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 04:03:51.368291 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.466667 | |
| I0330 04:03:51.368304 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.99375 (* 0.3 = 0.898124 loss) | |
| I0330 04:03:51.368319 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.838217 (* 0.3 = 0.251465 loss) | |
| I0330 04:03:51.368331 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.377778 | |
| I0330 04:03:51.368343 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 04:03:51.368355 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.555556 | |
| I0330 04:03:51.368368 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.57854 (* 1 = 2.57854 loss) | |
| I0330 04:03:51.368382 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.738439 (* 1 = 0.738439 loss) | |
| I0330 04:03:51.368394 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:03:51.368407 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00263084 | |
| I0330 04:03:51.368418 10583 sgd_solver.cpp:106] Iteration 41500, lr = 0.01 | |
| I0330 04:06:00.811022 10583 solver.cpp:229] Iteration 42000, loss = 5.00357 | |
| I0330 04:06:00.811151 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.183673 | |
| I0330 04:06:00.811172 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 04:06:00.811184 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.469388 | |
| I0330 04:06:00.811200 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5973 (* 0.3 = 0.779191 loss) | |
| I0330 04:06:00.811215 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.769538 (* 0.3 = 0.230861 loss) | |
| I0330 04:06:00.811228 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.244898 | |
| I0330 04:06:00.811240 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 04:06:00.811251 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.530612 | |
| I0330 04:06:00.811265 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.51769 (* 0.3 = 0.755307 loss) | |
| I0330 04:06:00.811280 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.739098 (* 0.3 = 0.22173 loss) | |
| I0330 04:06:00.811292 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.285714 | |
| I0330 04:06:00.811305 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 04:06:00.811316 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.77551 | |
| I0330 04:06:00.811331 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.03162 (* 1 = 2.03162 loss) | |
| I0330 04:06:00.811344 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.656336 (* 1 = 0.656336 loss) | |
| I0330 04:06:00.811357 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:06:00.811368 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00424339 | |
| I0330 04:06:00.811380 10583 sgd_solver.cpp:106] Iteration 42000, lr = 0.01 | |
| I0330 04:08:10.059761 10583 solver.cpp:229] Iteration 42500, loss = 4.9582 | |
| I0330 04:08:10.059898 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.24 | |
| I0330 04:08:10.059918 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 04:08:10.059931 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.34 | |
| I0330 04:08:10.059947 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82804 (* 0.3 = 0.848411 loss) | |
| I0330 04:08:10.059962 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.84524 (* 0.3 = 0.253572 loss) | |
| I0330 04:08:10.059975 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22 | |
| I0330 04:08:10.059988 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 04:08:10.059999 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.42 | |
| I0330 04:08:10.060014 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.60364 (* 0.3 = 0.781092 loss) | |
| I0330 04:08:10.060029 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.760787 (* 0.3 = 0.228236 loss) | |
| I0330 04:08:10.060040 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.36 | |
| I0330 04:08:10.060052 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0330 04:08:10.060065 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.52 | |
| I0330 04:08:10.060078 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.32193 (* 1 = 2.32193 loss) | |
| I0330 04:08:10.060092 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.678354 (* 1 = 0.678354 loss) | |
| I0330 04:08:10.060106 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:08:10.060117 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00665449 | |
| I0330 04:08:10.060130 10583 sgd_solver.cpp:106] Iteration 42500, lr = 0.01 | |
| I0330 04:10:19.263722 10583 solver.cpp:229] Iteration 43000, loss = 4.97144 | |
| I0330 04:10:19.263840 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0909091 | |
| I0330 04:10:19.263860 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 04:10:19.263875 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.227273 | |
| I0330 04:10:19.263890 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.02406 (* 0.3 = 0.907218 loss) | |
| I0330 04:10:19.263906 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.826755 (* 0.3 = 0.248026 loss) | |
| I0330 04:10:19.263917 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0909091 | |
| I0330 04:10:19.263931 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 04:10:19.263942 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.272727 | |
| I0330 04:10:19.263957 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.95841 (* 0.3 = 0.887522 loss) | |
| I0330 04:10:19.263970 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.807892 (* 0.3 = 0.242368 loss) | |
| I0330 04:10:19.263983 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.227273 | |
| I0330 04:10:19.263995 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 04:10:19.264008 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.409091 | |
| I0330 04:10:19.264021 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.57207 (* 1 = 2.57207 loss) | |
| I0330 04:10:19.264035 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.704621 (* 1 = 0.704621 loss) | |
| I0330 04:10:19.264048 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:10:19.264060 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00943398 | |
| I0330 04:10:19.264072 10583 sgd_solver.cpp:106] Iteration 43000, lr = 0.01 | |
| I0330 04:12:28.214684 10583 solver.cpp:229] Iteration 43500, loss = 4.92223 | |
| I0330 04:12:28.214839 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.169811 | |
| I0330 04:12:28.214871 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 04:12:28.214893 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.339623 | |
| I0330 04:12:28.214910 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.85597 (* 0.3 = 0.85679 loss) | |
| I0330 04:12:28.214926 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.914188 (* 0.3 = 0.274256 loss) | |
| I0330 04:12:28.214938 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.264151 | |
| I0330 04:12:28.214952 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 04:12:28.214964 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.433962 | |
| I0330 04:12:28.214995 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.57353 (* 0.3 = 0.77206 loss) | |
| I0330 04:12:28.215011 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.799656 (* 0.3 = 0.239897 loss) | |
| I0330 04:12:28.215024 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.396226 | |
| I0330 04:12:28.215036 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0330 04:12:28.215049 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.622642 | |
| I0330 04:12:28.215064 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.13778 (* 1 = 2.13778 loss) | |
| I0330 04:12:28.215077 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.672132 (* 1 = 0.672132 loss) | |
| I0330 04:12:28.215090 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:12:28.215102 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00757257 | |
| I0330 04:12:28.215117 10583 sgd_solver.cpp:106] Iteration 43500, lr = 0.01 | |
| I0330 04:14:37.221099 10583 solver.cpp:229] Iteration 44000, loss = 4.91491 | |
| I0330 04:14:37.221207 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.1875 | |
| I0330 04:14:37.221227 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 04:14:37.221240 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.333333 | |
| I0330 04:14:37.221257 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.91716 (* 0.3 = 0.875147 loss) | |
| I0330 04:14:37.221271 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.86323 (* 0.3 = 0.258969 loss) | |
| I0330 04:14:37.221284 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0330 04:14:37.221297 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 04:14:37.221308 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.395833 | |
| I0330 04:14:37.221323 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.75143 (* 0.3 = 0.82543 loss) | |
| I0330 04:14:37.221338 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.797651 (* 0.3 = 0.239295 loss) | |
| I0330 04:14:37.221349 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.375 | |
| I0330 04:14:37.221361 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0330 04:14:37.221374 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.604167 | |
| I0330 04:14:37.221388 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.3075 (* 1 = 2.3075 loss) | |
| I0330 04:14:37.221402 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.691192 (* 1 = 0.691192 loss) | |
| I0330 04:14:37.221415 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:14:37.221426 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0154304 | |
| I0330 04:14:37.221439 10583 sgd_solver.cpp:106] Iteration 44000, lr = 0.01 | |
| I0330 04:16:46.287758 10583 solver.cpp:229] Iteration 44500, loss = 4.84666 | |
| I0330 04:16:46.287895 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.108696 | |
| I0330 04:16:46.287915 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 04:16:46.287930 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.282609 | |
| I0330 04:16:46.287945 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.97944 (* 0.3 = 0.893831 loss) | |
| I0330 04:16:46.287962 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.841104 (* 0.3 = 0.252331 loss) | |
| I0330 04:16:46.287974 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0652174 | |
| I0330 04:16:46.287987 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0330 04:16:46.287999 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.23913 | |
| I0330 04:16:46.288013 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.88834 (* 0.3 = 0.866502 loss) | |
| I0330 04:16:46.288028 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.805058 (* 0.3 = 0.241518 loss) | |
| I0330 04:16:46.288041 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.369565 | |
| I0330 04:16:46.288053 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0330 04:16:46.288065 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.586957 | |
| I0330 04:16:46.288080 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.14217 (* 1 = 2.14217 loss) | |
| I0330 04:16:46.288094 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.592825 (* 1 = 0.592825 loss) | |
| I0330 04:16:46.288107 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:16:46.288120 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00365191 | |
| I0330 04:16:46.288131 10583 sgd_solver.cpp:106] Iteration 44500, lr = 0.01 | |
| I0330 04:18:55.895105 10583 solver.cpp:338] Iteration 45000, Testing net (#0) | |
| I0330 04:19:25.606330 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 04:19:25.606381 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 04:19:25.606398 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759728 | |
| I0330 04:19:25.606411 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 04:19:25.606428 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:19:25.606444 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:19:25.606456 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 04:19:25.606468 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759728 | |
| I0330 04:19:25.606480 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 04:19:25.606494 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:19:25.606508 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:19:25.606523 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 04:19:25.606534 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759728 | |
| I0330 04:19:25.606545 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 04:19:25.606559 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 04:19:25.606573 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 04:19:25.606586 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 04:19:25.606598 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 04:19:25.757486 10583 solver.cpp:229] Iteration 45000, loss = 4.81158 | |
| I0330 04:19:25.757532 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0330 04:19:25.757550 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 04:19:25.757562 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.413043 | |
| I0330 04:19:25.757578 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.6706 (* 0.3 = 0.801179 loss) | |
| I0330 04:19:25.757592 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.821073 (* 0.3 = 0.246322 loss) | |
| I0330 04:19:25.757606 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.217391 | |
| I0330 04:19:25.757618 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 04:19:25.757630 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.456522 | |
| I0330 04:19:25.757645 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.56459 (* 0.3 = 0.769378 loss) | |
| I0330 04:19:25.757663 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.740494 (* 0.3 = 0.222148 loss) | |
| I0330 04:19:25.757675 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.413043 | |
| I0330 04:19:25.757688 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0330 04:19:25.757699 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.673913 | |
| I0330 04:19:25.757714 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.81767 (* 1 = 1.81767 loss) | |
| I0330 04:19:25.757727 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.538109 (* 1 = 0.538109 loss) | |
| I0330 04:19:25.757740 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:19:25.757752 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00598219 | |
| I0330 04:19:25.757766 10583 sgd_solver.cpp:106] Iteration 45000, lr = 0.01 | |
| I0330 04:21:35.154389 10583 solver.cpp:229] Iteration 45500, loss = 4.78726 | |
| I0330 04:21:35.154561 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.219512 | |
| I0330 04:21:35.154582 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 04:21:35.154597 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.439024 | |
| I0330 04:21:35.154613 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.49661 (* 0.3 = 0.748982 loss) | |
| I0330 04:21:35.154628 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.699237 (* 0.3 = 0.209771 loss) | |
| I0330 04:21:35.154640 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.317073 | |
| I0330 04:21:35.154654 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 04:21:35.154665 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.585366 | |
| I0330 04:21:35.154680 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.10006 (* 0.3 = 0.630018 loss) | |
| I0330 04:21:35.154695 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.633045 (* 0.3 = 0.189914 loss) | |
| I0330 04:21:35.154707 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.536585 | |
| I0330 04:21:35.154719 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 04:21:35.154733 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804878 | |
| I0330 04:21:35.154747 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56765 (* 1 = 1.56765 loss) | |
| I0330 04:21:35.154762 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.449798 (* 1 = 0.449798 loss) | |
| I0330 04:21:35.154774 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 04:21:35.154788 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0568517 | |
| I0330 04:21:35.154800 10583 sgd_solver.cpp:106] Iteration 45500, lr = 0.01 | |
| I0330 04:23:45.312999 10583 solver.cpp:229] Iteration 46000, loss = 4.78131 | |
| I0330 04:23:45.313145 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.111111 | |
| I0330 04:23:45.313166 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0330 04:23:45.313179 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.333333 | |
| I0330 04:23:45.313196 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.81928 (* 0.3 = 0.845783 loss) | |
| I0330 04:23:45.313211 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.931436 (* 0.3 = 0.279431 loss) | |
| I0330 04:23:45.313225 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.277778 | |
| I0330 04:23:45.313244 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 04:23:45.313257 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.462963 | |
| I0330 04:23:45.313284 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.71648 (* 0.3 = 0.814943 loss) | |
| I0330 04:23:45.313308 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.91827 (* 0.3 = 0.275481 loss) | |
| I0330 04:23:45.313323 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.37037 | |
| I0330 04:23:45.313335 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 04:23:45.313349 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.574074 | |
| I0330 04:23:45.313362 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.30408 (* 1 = 2.30408 loss) | |
| I0330 04:23:45.313376 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.784759 (* 1 = 0.784759 loss) | |
| I0330 04:23:45.313390 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:23:45.313401 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00763799 | |
| I0330 04:23:45.313415 10583 sgd_solver.cpp:106] Iteration 46000, lr = 0.01 | |
| I0330 04:25:55.659373 10583 solver.cpp:229] Iteration 46500, loss = 4.77243 | |
| I0330 04:25:55.659546 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22449 | |
| I0330 04:25:55.659567 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 04:25:55.659580 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.469388 | |
| I0330 04:25:55.659598 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.83442 (* 0.3 = 0.850325 loss) | |
| I0330 04:25:55.659613 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.842595 (* 0.3 = 0.252778 loss) | |
| I0330 04:25:55.659626 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22449 | |
| I0330 04:25:55.659638 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 04:25:55.659651 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.530612 | |
| I0330 04:25:55.659664 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.63764 (* 0.3 = 0.791292 loss) | |
| I0330 04:25:55.659678 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.808817 (* 0.3 = 0.242645 loss) | |
| I0330 04:25:55.659692 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.44898 | |
| I0330 04:25:55.659703 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 04:25:55.659715 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795918 | |
| I0330 04:25:55.659729 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.79942 (* 1 = 1.79942 loss) | |
| I0330 04:25:55.659744 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.542469 (* 1 = 0.542469 loss) | |
| I0330 04:25:55.659755 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:25:55.659766 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00787726 | |
| I0330 04:25:55.659780 10583 sgd_solver.cpp:106] Iteration 46500, lr = 0.01 | |
| I0330 04:28:05.499296 10583 solver.cpp:229] Iteration 47000, loss = 4.76965 | |
| I0330 04:28:05.499496 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195652 | |
| I0330 04:28:05.499517 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 04:28:05.499532 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.434783 | |
| I0330 04:28:05.499550 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.603 (* 0.3 = 0.780901 loss) | |
| I0330 04:28:05.499564 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.750458 (* 0.3 = 0.225137 loss) | |
| I0330 04:28:05.499577 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.304348 | |
| I0330 04:28:05.499590 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 04:28:05.499603 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.586957 | |
| I0330 04:28:05.499616 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34092 (* 0.3 = 0.702276 loss) | |
| I0330 04:28:05.499631 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.657804 (* 0.3 = 0.197341 loss) | |
| I0330 04:28:05.499644 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.456522 | |
| I0330 04:28:05.499656 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 04:28:05.499668 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.717391 | |
| I0330 04:28:05.499682 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.96792 (* 1 = 1.96792 loss) | |
| I0330 04:28:05.499696 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.55042 (* 1 = 0.55042 loss) | |
| I0330 04:28:05.499708 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:28:05.499721 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0322735 | |
| I0330 04:28:05.499733 10583 sgd_solver.cpp:106] Iteration 47000, lr = 0.01 | |
| I0330 04:30:15.657130 10583 solver.cpp:229] Iteration 47500, loss = 4.74268 | |
| I0330 04:30:15.657297 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.235294 | |
| I0330 04:30:15.657318 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 04:30:15.657332 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.411765 | |
| I0330 04:30:15.657349 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65066 (* 0.3 = 0.795199 loss) | |
| I0330 04:30:15.657364 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.815157 (* 0.3 = 0.244547 loss) | |
| I0330 04:30:15.657377 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.27451 | |
| I0330 04:30:15.657389 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 04:30:15.657402 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.647059 | |
| I0330 04:30:15.657416 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.18444 (* 0.3 = 0.655332 loss) | |
| I0330 04:30:15.657430 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.682865 (* 0.3 = 0.204859 loss) | |
| I0330 04:30:15.657444 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.490196 | |
| I0330 04:30:15.657455 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 04:30:15.657467 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.705882 | |
| I0330 04:30:15.657482 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.74777 (* 1 = 1.74777 loss) | |
| I0330 04:30:15.657496 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.553356 (* 1 = 0.553356 loss) | |
| I0330 04:30:15.657510 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 04:30:15.657521 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0364795 | |
| I0330 04:30:15.657534 10583 sgd_solver.cpp:106] Iteration 47500, lr = 0.01 | |
| I0330 04:32:25.801224 10583 solver.cpp:229] Iteration 48000, loss = 4.63822 | |
| I0330 04:32:25.801394 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0330 04:32:25.801415 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 04:32:25.801429 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.38 | |
| I0330 04:32:25.801455 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64515 (* 0.3 = 0.793544 loss) | |
| I0330 04:32:25.801468 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.795857 (* 0.3 = 0.238757 loss) | |
| I0330 04:32:25.801481 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.42 | |
| I0330 04:32:25.801493 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 04:32:25.801506 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.62 | |
| I0330 04:32:25.801519 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.2858 (* 0.3 = 0.685739 loss) | |
| I0330 04:32:25.801533 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.709013 (* 0.3 = 0.212704 loss) | |
| I0330 04:32:25.801545 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.56 | |
| I0330 04:32:25.801558 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 04:32:25.801569 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 04:32:25.801584 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56 (* 1 = 1.56 loss) | |
| I0330 04:32:25.801597 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.511048 (* 1 = 0.511048 loss) | |
| I0330 04:32:25.801610 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 04:32:25.801622 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0251785 | |
| I0330 04:32:25.801635 10583 sgd_solver.cpp:106] Iteration 48000, lr = 0.01 | |
| I0330 04:34:35.823717 10583 solver.cpp:229] Iteration 48500, loss = 4.63795 | |
| I0330 04:34:35.823881 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145455 | |
| I0330 04:34:35.823902 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0330 04:34:35.823915 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.345455 | |
| I0330 04:34:35.823932 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.91496 (* 0.3 = 0.874487 loss) | |
| I0330 04:34:35.823947 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.959407 (* 0.3 = 0.287822 loss) | |
| I0330 04:34:35.823961 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.272727 | |
| I0330 04:34:35.823973 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 04:34:35.823985 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.436364 | |
| I0330 04:34:35.824000 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98009 (* 0.3 = 0.894028 loss) | |
| I0330 04:34:35.824015 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.966864 (* 0.3 = 0.290059 loss) | |
| I0330 04:34:35.824028 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.327273 | |
| I0330 04:34:35.824040 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0330 04:34:35.824053 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.636364 | |
| I0330 04:34:35.824067 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.24791 (* 1 = 2.24791 loss) | |
| I0330 04:34:35.824081 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.764165 (* 1 = 0.764165 loss) | |
| I0330 04:34:35.824095 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:34:35.824106 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0129541 | |
| I0330 04:34:35.824120 10583 sgd_solver.cpp:106] Iteration 48500, lr = 0.01 | |
| I0330 04:36:45.743513 10583 solver.cpp:229] Iteration 49000, loss = 4.59939 | |
| I0330 04:36:45.743688 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.194444 | |
| I0330 04:36:45.743710 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 04:36:45.743723 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.472222 | |
| I0330 04:36:45.743739 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.47386 (* 0.3 = 0.742157 loss) | |
| I0330 04:36:45.743762 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.700061 (* 0.3 = 0.210018 loss) | |
| I0330 04:36:45.743774 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.277778 | |
| I0330 04:36:45.743788 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 04:36:45.743799 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 04:36:45.743813 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.44465 (* 0.3 = 0.733396 loss) | |
| I0330 04:36:45.743827 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.646221 (* 0.3 = 0.193866 loss) | |
| I0330 04:36:45.743840 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.361111 | |
| I0330 04:36:45.743852 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 04:36:45.743865 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.555556 | |
| I0330 04:36:45.743880 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.0345 (* 1 = 2.0345 loss) | |
| I0330 04:36:45.743893 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.540151 (* 1 = 0.540151 loss) | |
| I0330 04:36:45.743906 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:36:45.743918 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00575829 | |
| I0330 04:36:45.743932 10583 sgd_solver.cpp:106] Iteration 49000, lr = 0.01 | |
| I0330 04:38:54.982831 10583 solver.cpp:229] Iteration 49500, loss = 4.59774 | |
| I0330 04:38:54.982946 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.134615 | |
| I0330 04:38:54.982966 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 04:38:54.982980 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.326923 | |
| I0330 04:38:54.982995 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.8728 (* 0.3 = 0.861839 loss) | |
| I0330 04:38:54.983009 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.90953 (* 0.3 = 0.272859 loss) | |
| I0330 04:38:54.983022 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.153846 | |
| I0330 04:38:54.983034 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 04:38:54.983047 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.346154 | |
| I0330 04:38:54.983060 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.88107 (* 0.3 = 0.864322 loss) | |
| I0330 04:38:54.983088 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.891876 (* 0.3 = 0.267563 loss) | |
| I0330 04:38:54.983101 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.288462 | |
| I0330 04:38:54.983114 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0330 04:38:54.983125 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.538462 | |
| I0330 04:38:54.983140 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.40043 (* 1 = 2.40043 loss) | |
| I0330 04:38:54.983155 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.762098 (* 1 = 0.762098 loss) | |
| I0330 04:38:54.983170 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:38:54.983181 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0138883 | |
| I0330 04:38:54.983194 10583 sgd_solver.cpp:106] Iteration 49500, lr = 0.01 | |
| I0330 04:41:04.088913 10583 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_50000.caffemodel | |
| I0330 04:41:04.455188 10583 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_50000.solverstate | |
| I0330 04:41:04.619309 10583 solver.cpp:338] Iteration 50000, Testing net (#0) | |
| I0330 04:41:34.495957 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 04:41:34.496071 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 04:41:34.496090 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760001 | |
| I0330 04:41:34.496104 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 04:41:34.496121 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:41:34.496139 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:41:34.496150 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 04:41:34.496165 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760001 | |
| I0330 04:41:34.496177 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 04:41:34.496191 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:41:34.496206 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 04:41:34.496219 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 04:41:34.496232 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760001 | |
| I0330 04:41:34.496243 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 04:41:34.496258 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 04:41:34.496271 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 04:41:34.496284 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 04:41:34.496295 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 04:41:34.647802 10583 solver.cpp:229] Iteration 50000, loss = 4.60658 | |
| I0330 04:41:34.647857 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0330 04:41:34.647874 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 04:41:34.647897 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.36 | |
| I0330 04:41:34.647912 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.1121 (* 0.3 = 0.933629 loss) | |
| I0330 04:41:34.647927 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.952424 (* 0.3 = 0.285727 loss) | |
| I0330 04:41:34.647938 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.14 | |
| I0330 04:41:34.647951 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 04:41:34.647967 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.36 | |
| I0330 04:41:34.647990 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.42374 (* 0.3 = 1.02712 loss) | |
| I0330 04:41:34.648005 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.0336 (* 0.3 = 0.310079 loss) | |
| I0330 04:41:34.648016 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.28 | |
| I0330 04:41:34.648028 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0330 04:41:34.648041 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.52 | |
| I0330 04:41:34.648062 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.55187 (* 1 = 2.55187 loss) | |
| I0330 04:41:34.648077 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.78914 (* 1 = 0.78914 loss) | |
| I0330 04:41:34.648088 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:41:34.648100 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0085794 | |
| I0330 04:41:34.648113 10583 sgd_solver.cpp:106] Iteration 50000, lr = 0.01 | |
| I0330 04:43:43.815860 10583 solver.cpp:229] Iteration 50500, loss = 4.54871 | |
| I0330 04:43:43.816009 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.142857 | |
| I0330 04:43:43.816030 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 04:43:43.816042 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.333333 | |
| I0330 04:43:43.816058 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.11581 (* 0.3 = 0.934742 loss) | |
| I0330 04:43:43.816074 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.930558 (* 0.3 = 0.279167 loss) | |
| I0330 04:43:43.816095 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.238095 | |
| I0330 04:43:43.816107 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 04:43:43.816119 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.547619 | |
| I0330 04:43:43.816133 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.66251 (* 0.3 = 0.798754 loss) | |
| I0330 04:43:43.816148 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.79119 (* 0.3 = 0.237357 loss) | |
| I0330 04:43:43.816162 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.404762 | |
| I0330 04:43:43.816175 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 04:43:43.816187 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0330 04:43:43.816201 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.29193 (* 1 = 2.29193 loss) | |
| I0330 04:43:43.816215 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.681002 (* 1 = 0.681002 loss) | |
| I0330 04:43:43.816228 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:43:43.816241 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0543882 | |
| I0330 04:43:43.816253 10583 sgd_solver.cpp:106] Iteration 50500, lr = 0.01 | |
| I0330 04:45:53.023545 10583 solver.cpp:229] Iteration 51000, loss = 4.51806 | |
| I0330 04:45:53.023674 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.261905 | |
| I0330 04:45:53.023694 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 04:45:53.023707 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.428571 | |
| I0330 04:45:53.023725 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76749 (* 0.3 = 0.830248 loss) | |
| I0330 04:45:53.023739 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.722606 (* 0.3 = 0.216782 loss) | |
| I0330 04:45:53.023751 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.261905 | |
| I0330 04:45:53.023763 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 04:45:53.023775 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 04:45:53.023789 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.57006 (* 0.3 = 0.771017 loss) | |
| I0330 04:45:53.023804 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.705047 (* 0.3 = 0.211514 loss) | |
| I0330 04:45:53.023816 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.380952 | |
| I0330 04:45:53.023828 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 04:45:53.023840 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.642857 | |
| I0330 04:45:53.023854 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.01901 (* 1 = 2.01901 loss) | |
| I0330 04:45:53.023869 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.570426 (* 1 = 0.570426 loss) | |
| I0330 04:45:53.023880 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:45:53.023892 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0246562 | |
| I0330 04:45:53.023905 10583 sgd_solver.cpp:106] Iteration 51000, lr = 0.01 | |
| I0330 04:48:02.122880 10583 solver.cpp:229] Iteration 51500, loss = 4.53836 | |
| I0330 04:48:02.123024 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0330 04:48:02.123046 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 04:48:02.123059 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0330 04:48:02.123075 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39302 (* 0.3 = 0.717907 loss) | |
| I0330 04:48:02.123090 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.689873 (* 0.3 = 0.206962 loss) | |
| I0330 04:48:02.123102 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.282609 | |
| I0330 04:48:02.123122 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 04:48:02.123134 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.695652 | |
| I0330 04:48:02.123147 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.17552 (* 0.3 = 0.652656 loss) | |
| I0330 04:48:02.123162 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.614244 (* 0.3 = 0.184273 loss) | |
| I0330 04:48:02.123174 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.521739 | |
| I0330 04:48:02.123186 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 04:48:02.123198 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826087 | |
| I0330 04:48:02.123213 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.66111 (* 1 = 1.66111 loss) | |
| I0330 04:48:02.123226 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.460891 (* 1 = 0.460891 loss) | |
| I0330 04:48:02.123239 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:48:02.123250 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0354035 | |
| I0330 04:48:02.123262 10583 sgd_solver.cpp:106] Iteration 51500, lr = 0.01 | |
| I0330 04:50:11.266772 10583 solver.cpp:229] Iteration 52000, loss = 4.46322 | |
| I0330 04:50:11.266888 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.254902 | |
| I0330 04:50:11.266908 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 04:50:11.266921 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.470588 | |
| I0330 04:50:11.266938 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76364 (* 0.3 = 0.829091 loss) | |
| I0330 04:50:11.266952 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.842938 (* 0.3 = 0.252881 loss) | |
| I0330 04:50:11.266965 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.294118 | |
| I0330 04:50:11.266978 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 04:50:11.266989 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.588235 | |
| I0330 04:50:11.267004 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34788 (* 0.3 = 0.704363 loss) | |
| I0330 04:50:11.267031 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.748904 (* 0.3 = 0.224671 loss) | |
| I0330 04:50:11.267045 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.470588 | |
| I0330 04:50:11.267057 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 04:50:11.267069 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0330 04:50:11.267083 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.9126 (* 1 = 1.9126 loss) | |
| I0330 04:50:11.267097 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.605952 (* 1 = 0.605952 loss) | |
| I0330 04:50:11.267110 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 04:50:11.267122 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0921056 | |
| I0330 04:50:11.267135 10583 sgd_solver.cpp:106] Iteration 52000, lr = 0.01 | |
| I0330 04:52:20.460880 10583 solver.cpp:229] Iteration 52500, loss = 4.50396 | |
| I0330 04:52:20.461024 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.209302 | |
| I0330 04:52:20.461045 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 04:52:20.461057 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.348837 | |
| I0330 04:52:20.461073 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.17862 (* 0.3 = 0.953587 loss) | |
| I0330 04:52:20.461088 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.924783 (* 0.3 = 0.277435 loss) | |
| I0330 04:52:20.461107 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.348837 | |
| I0330 04:52:20.461119 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 04:52:20.461132 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.581395 | |
| I0330 04:52:20.461145 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.56973 (* 0.3 = 0.770919 loss) | |
| I0330 04:52:20.461163 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.753303 (* 0.3 = 0.225991 loss) | |
| I0330 04:52:20.461175 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.55814 | |
| I0330 04:52:20.461187 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 04:52:20.461199 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72093 | |
| I0330 04:52:20.461213 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.94248 (* 1 = 1.94248 loss) | |
| I0330 04:52:20.461228 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.607057 (* 1 = 0.607057 loss) | |
| I0330 04:52:20.461241 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:52:20.461253 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0132669 | |
| I0330 04:52:20.461266 10583 sgd_solver.cpp:106] Iteration 52500, lr = 0.01 | |
| I0330 04:54:29.322945 10583 solver.cpp:229] Iteration 53000, loss = 4.43759 | |
| I0330 04:54:29.323063 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.153846 | |
| I0330 04:54:29.323083 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 04:54:29.323097 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.326923 | |
| I0330 04:54:29.323113 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.80781 (* 0.3 = 0.842344 loss) | |
| I0330 04:54:29.323128 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.86482 (* 0.3 = 0.259446 loss) | |
| I0330 04:54:29.323140 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.25 | |
| I0330 04:54:29.323153 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 04:54:29.323168 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.480769 | |
| I0330 04:54:29.323182 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.57276 (* 0.3 = 0.771829 loss) | |
| I0330 04:54:29.323196 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.824539 (* 0.3 = 0.247362 loss) | |
| I0330 04:54:29.323210 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.423077 | |
| I0330 04:54:29.323221 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 04:54:29.323232 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.711538 | |
| I0330 04:54:29.323246 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.88797 (* 1 = 1.88797 loss) | |
| I0330 04:54:29.323261 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.601905 (* 1 = 0.601905 loss) | |
| I0330 04:54:29.323273 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:54:29.323284 10583 solver.cpp:245] Train net output #16: total_confidence = 0.00906129 | |
| I0330 04:54:29.323297 10583 sgd_solver.cpp:106] Iteration 53000, lr = 0.01 | |
| I0330 04:56:38.592864 10583 solver.cpp:229] Iteration 53500, loss = 4.41666 | |
| I0330 04:56:38.593014 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 04:56:38.593034 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 04:56:38.593047 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 04:56:38.593065 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.37246 (* 0.3 = 0.711739 loss) | |
| I0330 04:56:38.593088 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.701026 (* 0.3 = 0.210308 loss) | |
| I0330 04:56:38.593101 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0330 04:56:38.593113 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 04:56:38.593124 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 04:56:38.593139 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.30861 (* 0.3 = 0.692582 loss) | |
| I0330 04:56:38.593154 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.673684 (* 0.3 = 0.202105 loss) | |
| I0330 04:56:38.593169 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.541667 | |
| I0330 04:56:38.593183 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 04:56:38.593194 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.791667 | |
| I0330 04:56:38.593209 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.50383 (* 1 = 1.50383 loss) | |
| I0330 04:56:38.593222 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.444779 (* 1 = 0.444779 loss) | |
| I0330 04:56:38.593235 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 04:56:38.593246 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0436973 | |
| I0330 04:56:38.593260 10583 sgd_solver.cpp:106] Iteration 53500, lr = 0.01 | |
| I0330 04:58:48.019202 10583 solver.cpp:229] Iteration 54000, loss = 4.4871 | |
| I0330 04:58:48.019327 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.238095 | |
| I0330 04:58:48.019347 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 04:58:48.019361 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.47619 | |
| I0330 04:58:48.019377 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5864 (* 0.3 = 0.77592 loss) | |
| I0330 04:58:48.019392 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.766066 (* 0.3 = 0.22982 loss) | |
| I0330 04:58:48.019404 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.357143 | |
| I0330 04:58:48.019417 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 04:58:48.019428 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 04:58:48.019443 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.30998 (* 0.3 = 0.692993 loss) | |
| I0330 04:58:48.019457 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.713917 (* 0.3 = 0.214175 loss) | |
| I0330 04:58:48.019470 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.428571 | |
| I0330 04:58:48.019482 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 04:58:48.019495 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.761905 | |
| I0330 04:58:48.019508 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.86332 (* 1 = 1.86332 loss) | |
| I0330 04:58:48.019523 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.551284 (* 1 = 0.551284 loss) | |
| I0330 04:58:48.019536 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 04:58:48.019547 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0260813 | |
| I0330 04:58:48.019561 10583 sgd_solver.cpp:106] Iteration 54000, lr = 0.01 | |
| I0330 05:00:57.192627 10583 solver.cpp:229] Iteration 54500, loss = 4.34363 | |
| I0330 05:00:57.192759 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.230769 | |
| I0330 05:00:57.192778 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 05:00:57.192791 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.346154 | |
| I0330 05:00:57.192808 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.92507 (* 0.3 = 0.87752 loss) | |
| I0330 05:00:57.192823 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.892615 (* 0.3 = 0.267784 loss) | |
| I0330 05:00:57.192842 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.192308 | |
| I0330 05:00:57.192855 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 05:00:57.192867 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.442308 | |
| I0330 05:00:57.192881 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85631 (* 0.3 = 0.856894 loss) | |
| I0330 05:00:57.192895 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.862943 (* 0.3 = 0.258883 loss) | |
| I0330 05:00:57.192908 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.346154 | |
| I0330 05:00:57.192919 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0330 05:00:57.192931 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.519231 | |
| I0330 05:00:57.192945 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.57114 (* 1 = 2.57114 loss) | |
| I0330 05:00:57.192960 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.790255 (* 1 = 0.790255 loss) | |
| I0330 05:00:57.192972 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:00:57.192984 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0148263 | |
| I0330 05:00:57.192996 10583 sgd_solver.cpp:106] Iteration 54500, lr = 0.01 | |
| I0330 05:03:06.224696 10583 solver.cpp:338] Iteration 55000, Testing net (#0) | |
| I0330 05:03:36.022356 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 05:03:36.022421 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 05:03:36.022439 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759183 | |
| I0330 05:03:36.022451 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 05:03:36.022469 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:03:36.022485 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:03:36.022497 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 05:03:36.022510 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759183 | |
| I0330 05:03:36.022521 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 05:03:36.022536 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:03:36.022550 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:03:36.022562 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 05:03:36.022574 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759183 | |
| I0330 05:03:36.022585 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 05:03:36.022599 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:03:36.022614 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:03:36.022626 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 05:03:36.022639 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 05:03:36.173629 10583 solver.cpp:229] Iteration 55000, loss = 4.38047 | |
| I0330 05:03:36.173678 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0330 05:03:36.173694 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 05:03:36.173707 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.456522 | |
| I0330 05:03:36.173732 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.80314 (* 0.3 = 0.840941 loss) | |
| I0330 05:03:36.173746 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.844077 (* 0.3 = 0.253223 loss) | |
| I0330 05:03:36.173758 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.173913 | |
| I0330 05:03:36.173770 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 05:03:36.173782 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.434783 | |
| I0330 05:03:36.173796 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.80522 (* 0.3 = 0.841567 loss) | |
| I0330 05:03:36.173810 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.832392 (* 0.3 = 0.249718 loss) | |
| I0330 05:03:36.173822 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.326087 | |
| I0330 05:03:36.173835 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 05:03:36.173846 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.5 | |
| I0330 05:03:36.173861 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.31286 (* 1 = 2.31286 loss) | |
| I0330 05:03:36.173874 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.720292 (* 1 = 0.720292 loss) | |
| I0330 05:03:36.173887 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:03:36.173899 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0325221 | |
| I0330 05:03:36.173912 10583 sgd_solver.cpp:106] Iteration 55000, lr = 0.01 | |
| I0330 05:05:45.209002 10583 solver.cpp:229] Iteration 55500, loss = 4.31342 | |
| I0330 05:05:45.209134 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0909091 | |
| I0330 05:05:45.209154 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 05:05:45.209167 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.340909 | |
| I0330 05:05:45.209184 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.02964 (* 0.3 = 0.908893 loss) | |
| I0330 05:05:45.209206 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.857141 (* 0.3 = 0.257142 loss) | |
| I0330 05:05:45.209218 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.181818 | |
| I0330 05:05:45.209231 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 05:05:45.209244 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.431818 | |
| I0330 05:05:45.209256 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.90456 (* 0.3 = 0.871368 loss) | |
| I0330 05:05:45.209271 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.830783 (* 0.3 = 0.249235 loss) | |
| I0330 05:05:45.209285 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.340909 | |
| I0330 05:05:45.209296 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 05:05:45.209308 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.545455 | |
| I0330 05:05:45.209322 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.37489 (* 1 = 2.37489 loss) | |
| I0330 05:05:45.209336 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.644423 (* 1 = 0.644423 loss) | |
| I0330 05:05:45.209348 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:05:45.209359 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0490962 | |
| I0330 05:05:45.209372 10583 sgd_solver.cpp:106] Iteration 55500, lr = 0.01 | |
| I0330 05:07:54.159768 10583 solver.cpp:229] Iteration 56000, loss = 4.31769 | |
| I0330 05:07:54.159915 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.192982 | |
| I0330 05:07:54.159936 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0330 05:07:54.159950 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.421053 | |
| I0330 05:07:54.159965 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.74296 (* 0.3 = 0.822889 loss) | |
| I0330 05:07:54.159987 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.926537 (* 0.3 = 0.277961 loss) | |
| I0330 05:07:54.160001 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.210526 | |
| I0330 05:07:54.160012 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0330 05:07:54.160024 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.45614 | |
| I0330 05:07:54.160038 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.63918 (* 0.3 = 0.791753 loss) | |
| I0330 05:07:54.160053 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.871712 (* 0.3 = 0.261514 loss) | |
| I0330 05:07:54.160065 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.368421 | |
| I0330 05:07:54.160078 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 05:07:54.160089 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.631579 | |
| I0330 05:07:54.160104 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.01004 (* 1 = 2.01004 loss) | |
| I0330 05:07:54.160117 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.697819 (* 1 = 0.697819 loss) | |
| I0330 05:07:54.160130 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:07:54.160141 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0296657 | |
| I0330 05:07:54.160153 10583 sgd_solver.cpp:106] Iteration 56000, lr = 0.01 | |
| I0330 05:10:03.205525 10583 solver.cpp:229] Iteration 56500, loss = 4.33359 | |
| I0330 05:10:03.205628 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0330 05:10:03.205647 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 05:10:03.205662 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.44 | |
| I0330 05:10:03.205678 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65635 (* 0.3 = 0.796906 loss) | |
| I0330 05:10:03.205694 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.801947 (* 0.3 = 0.240584 loss) | |
| I0330 05:10:03.205708 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.18 | |
| I0330 05:10:03.205721 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 05:10:03.205734 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 05:10:03.205747 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.54216 (* 0.3 = 0.762647 loss) | |
| I0330 05:10:03.205761 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.764259 (* 0.3 = 0.229278 loss) | |
| I0330 05:10:03.205775 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.4 | |
| I0330 05:10:03.205786 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 05:10:03.205798 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.66 | |
| I0330 05:10:03.205812 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.93799 (* 1 = 1.93799 loss) | |
| I0330 05:10:03.205827 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.574331 (* 1 = 0.574331 loss) | |
| I0330 05:10:03.205839 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:10:03.205852 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0485685 | |
| I0330 05:10:03.205863 10583 sgd_solver.cpp:106] Iteration 56500, lr = 0.01 | |
| I0330 05:12:12.322959 10583 solver.cpp:229] Iteration 57000, loss = 4.32053 | |
| I0330 05:12:12.323143 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 05:12:12.323163 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 05:12:12.323178 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.457143 | |
| I0330 05:12:12.323194 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70098 (* 0.3 = 0.810293 loss) | |
| I0330 05:12:12.323209 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.779315 (* 0.3 = 0.233795 loss) | |
| I0330 05:12:12.323221 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.285714 | |
| I0330 05:12:12.323233 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 05:12:12.323246 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.628571 | |
| I0330 05:12:12.323259 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.38481 (* 0.3 = 0.715444 loss) | |
| I0330 05:12:12.323273 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.783132 (* 0.3 = 0.23494 loss) | |
| I0330 05:12:12.323287 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.485714 | |
| I0330 05:12:12.323298 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 05:12:12.323310 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.6 | |
| I0330 05:12:12.323325 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.85749 (* 1 = 1.85749 loss) | |
| I0330 05:12:12.323341 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.642947 (* 1 = 0.642947 loss) | |
| I0330 05:12:12.323354 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:12:12.323365 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0704309 | |
| I0330 05:12:12.323379 10583 sgd_solver.cpp:106] Iteration 57000, lr = 0.01 | |
| I0330 05:14:21.518391 10583 solver.cpp:229] Iteration 57500, loss = 4.21214 | |
| I0330 05:14:21.518506 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.23913 | |
| I0330 05:14:21.518525 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 05:14:21.518538 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.413043 | |
| I0330 05:14:21.518555 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.8991 (* 0.3 = 0.869731 loss) | |
| I0330 05:14:21.518570 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.924793 (* 0.3 = 0.277438 loss) | |
| I0330 05:14:21.518584 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.26087 | |
| I0330 05:14:21.518595 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 05:14:21.518607 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0330 05:14:21.518621 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.73829 (* 0.3 = 0.821487 loss) | |
| I0330 05:14:21.518635 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.900174 (* 0.3 = 0.270052 loss) | |
| I0330 05:14:21.518647 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.369565 | |
| I0330 05:14:21.518659 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0330 05:14:21.518671 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.652174 | |
| I0330 05:14:21.518685 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.30747 (* 1 = 2.30747 loss) | |
| I0330 05:14:21.518699 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.760733 (* 1 = 0.760733 loss) | |
| I0330 05:14:21.518712 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:14:21.518723 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0209336 | |
| I0330 05:14:21.518735 10583 sgd_solver.cpp:106] Iteration 57500, lr = 0.01 | |
| I0330 05:16:30.660332 10583 solver.cpp:229] Iteration 58000, loss = 4.27012 | |
| I0330 05:16:30.660473 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.166667 | |
| I0330 05:16:30.660495 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 05:16:30.660506 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.395833 | |
| I0330 05:16:30.660523 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.0572 (* 0.3 = 0.91716 loss) | |
| I0330 05:16:30.660538 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.877125 (* 0.3 = 0.263137 loss) | |
| I0330 05:16:30.660550 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.166667 | |
| I0330 05:16:30.660562 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 05:16:30.660574 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.375 | |
| I0330 05:16:30.660588 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85086 (* 0.3 = 0.855259 loss) | |
| I0330 05:16:30.660603 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.812138 (* 0.3 = 0.243641 loss) | |
| I0330 05:16:30.660614 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.479167 | |
| I0330 05:16:30.660626 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 05:16:30.660639 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.625 | |
| I0330 05:16:30.660652 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.26334 (* 1 = 2.26334 loss) | |
| I0330 05:16:30.660667 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.635078 (* 1 = 0.635078 loss) | |
| I0330 05:16:30.660679 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:16:30.660691 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0736971 | |
| I0330 05:16:30.660703 10583 sgd_solver.cpp:106] Iteration 58000, lr = 0.01 | |
| I0330 05:18:40.086160 10583 solver.cpp:229] Iteration 58500, loss = 4.22209 | |
| I0330 05:18:40.086290 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0330 05:18:40.086310 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 05:18:40.086324 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 05:18:40.086340 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.44473 (* 0.3 = 0.733419 loss) | |
| I0330 05:18:40.086356 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.741273 (* 0.3 = 0.222382 loss) | |
| I0330 05:18:40.086369 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.347826 | |
| I0330 05:18:40.086381 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 05:18:40.086392 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.608696 | |
| I0330 05:18:40.086406 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.25634 (* 0.3 = 0.676903 loss) | |
| I0330 05:18:40.086421 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.653571 (* 0.3 = 0.196071 loss) | |
| I0330 05:18:40.086433 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.456522 | |
| I0330 05:18:40.086446 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 05:18:40.086457 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.782609 | |
| I0330 05:18:40.086472 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.57411 (* 1 = 1.57411 loss) | |
| I0330 05:18:40.086486 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.44262 (* 1 = 0.44262 loss) | |
| I0330 05:18:40.086498 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:18:40.086510 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0951107 | |
| I0330 05:18:40.086524 10583 sgd_solver.cpp:106] Iteration 58500, lr = 0.01 | |
| I0330 05:20:49.464179 10583 solver.cpp:229] Iteration 59000, loss = 4.24651 | |
| I0330 05:20:49.464331 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.264151 | |
| I0330 05:20:49.464352 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 05:20:49.464373 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.45283 | |
| I0330 05:20:49.464390 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.77356 (* 0.3 = 0.832067 loss) | |
| I0330 05:20:49.464406 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.894009 (* 0.3 = 0.268203 loss) | |
| I0330 05:20:49.464418 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.264151 | |
| I0330 05:20:49.464431 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0330 05:20:49.464442 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.528302 | |
| I0330 05:20:49.464457 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.52903 (* 0.3 = 0.75871 loss) | |
| I0330 05:20:49.464470 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.905977 (* 0.3 = 0.271793 loss) | |
| I0330 05:20:49.464483 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.45283 | |
| I0330 05:20:49.464495 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0330 05:20:49.464506 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.603774 | |
| I0330 05:20:49.464521 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87214 (* 1 = 1.87214 loss) | |
| I0330 05:20:49.464535 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.664146 (* 1 = 0.664146 loss) | |
| I0330 05:20:49.464547 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:20:49.464560 10583 solver.cpp:245] Train net output #16: total_confidence = 0.037612 | |
| I0330 05:20:49.464573 10583 sgd_solver.cpp:106] Iteration 59000, lr = 0.01 | |
| I0330 05:22:58.927969 10583 solver.cpp:229] Iteration 59500, loss = 4.19278 | |
| I0330 05:22:58.928146 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.225 | |
| I0330 05:22:58.928170 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 05:22:58.928184 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.525 | |
| I0330 05:22:58.928201 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.7421 (* 0.3 = 0.822631 loss) | |
| I0330 05:22:58.928217 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.751913 (* 0.3 = 0.225574 loss) | |
| I0330 05:22:58.928239 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.35 | |
| I0330 05:22:58.928256 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 05:22:58.928268 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0330 05:22:58.928287 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.24426 (* 0.3 = 0.673279 loss) | |
| I0330 05:22:58.928303 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.598152 (* 0.3 = 0.179446 loss) | |
| I0330 05:22:58.928314 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.65 | |
| I0330 05:22:58.928326 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0330 05:22:58.928347 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 05:22:58.928362 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26996 (* 1 = 1.26996 loss) | |
| I0330 05:22:58.928375 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.435762 (* 1 = 0.435762 loss) | |
| I0330 05:22:58.928388 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:22:58.928400 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0742676 | |
| I0330 05:22:58.928413 10583 sgd_solver.cpp:106] Iteration 59500, lr = 0.01 | |
| I0330 05:25:08.032074 10583 solver.cpp:338] Iteration 60000, Testing net (#0) | |
| I0330 05:25:37.817245 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 05:25:37.817294 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 05:25:37.817317 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760137 | |
| I0330 05:25:37.817329 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 05:25:37.817347 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:25:37.817361 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:25:37.817381 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 05:25:37.817394 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760137 | |
| I0330 05:25:37.817404 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 05:25:37.817419 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:25:37.817433 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:25:37.817445 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 05:25:37.817457 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760137 | |
| I0330 05:25:37.817469 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 05:25:37.817483 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:25:37.817498 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:25:37.817510 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 05:25:37.817522 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 05:25:37.969233 10583 solver.cpp:229] Iteration 60000, loss = 4.11893 | |
| I0330 05:25:37.969271 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0330 05:25:37.969287 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 05:25:37.969300 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.521739 | |
| I0330 05:25:37.969316 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.45014 (* 0.3 = 0.735043 loss) | |
| I0330 05:25:37.969331 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.715842 (* 0.3 = 0.214753 loss) | |
| I0330 05:25:37.969343 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.282609 | |
| I0330 05:25:37.969355 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 05:25:37.969367 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.543478 | |
| I0330 05:25:37.969382 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34467 (* 0.3 = 0.703401 loss) | |
| I0330 05:25:37.969395 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.678267 (* 0.3 = 0.20348 loss) | |
| I0330 05:25:37.969408 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.521739 | |
| I0330 05:25:37.969420 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 05:25:37.969432 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0330 05:25:37.969446 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28905 (* 1 = 1.28905 loss) | |
| I0330 05:25:37.969460 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.384043 (* 1 = 0.384043 loss) | |
| I0330 05:25:37.969472 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:25:37.969485 10583 solver.cpp:245] Train net output #16: total_confidence = 0.106595 | |
| I0330 05:25:37.969498 10583 sgd_solver.cpp:106] Iteration 60000, lr = 0.01 | |
| I0330 05:27:46.969597 10583 solver.cpp:229] Iteration 60500, loss = 4.16815 | |
| I0330 05:27:46.969743 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 05:27:46.969764 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 05:27:46.969777 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.428571 | |
| I0330 05:27:46.969802 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.54449 (* 0.3 = 0.763347 loss) | |
| I0330 05:27:46.969817 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.739826 (* 0.3 = 0.221948 loss) | |
| I0330 05:27:46.969830 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.285714 | |
| I0330 05:27:46.969841 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 05:27:46.969853 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.452381 | |
| I0330 05:27:46.969867 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.53504 (* 0.3 = 0.760513 loss) | |
| I0330 05:27:46.969882 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.694483 (* 0.3 = 0.208345 loss) | |
| I0330 05:27:46.969893 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 05:27:46.969905 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 05:27:46.969918 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.690476 | |
| I0330 05:27:46.969931 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.79945 (* 1 = 1.79945 loss) | |
| I0330 05:27:46.969945 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.532426 (* 1 = 0.532426 loss) | |
| I0330 05:27:46.969959 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:27:46.969969 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0176774 | |
| I0330 05:27:46.969982 10583 sgd_solver.cpp:106] Iteration 60500, lr = 0.01 | |
| I0330 05:29:56.041877 10583 solver.cpp:229] Iteration 61000, loss = 4.13652 | |
| I0330 05:29:56.041991 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0330 05:29:56.042012 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 05:29:56.042026 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.488889 | |
| I0330 05:29:56.042042 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41587 (* 0.3 = 0.724761 loss) | |
| I0330 05:29:56.042057 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650217 (* 0.3 = 0.195065 loss) | |
| I0330 05:29:56.042069 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377778 | |
| I0330 05:29:56.042081 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 05:29:56.042093 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0330 05:29:56.042106 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12431 (* 0.3 = 0.637292 loss) | |
| I0330 05:29:56.042121 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.569287 (* 0.3 = 0.170786 loss) | |
| I0330 05:29:56.042134 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.555556 | |
| I0330 05:29:56.042146 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 05:29:56.042160 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0330 05:29:56.042176 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.41813 (* 1 = 1.41813 loss) | |
| I0330 05:29:56.042189 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.376319 (* 1 = 0.376319 loss) | |
| I0330 05:29:56.042201 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:29:56.042213 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0387987 | |
| I0330 05:29:56.042225 10583 sgd_solver.cpp:106] Iteration 61000, lr = 0.01 | |
| I0330 05:32:05.311354 10583 solver.cpp:229] Iteration 61500, loss = 4.17124 | |
| I0330 05:32:05.311517 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.188679 | |
| I0330 05:32:05.311537 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 05:32:05.311550 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.415094 | |
| I0330 05:32:05.311575 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.74829 (* 0.3 = 0.824486 loss) | |
| I0330 05:32:05.311590 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.868104 (* 0.3 = 0.260431 loss) | |
| I0330 05:32:05.311604 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.283019 | |
| I0330 05:32:05.311616 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 05:32:05.311628 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.566038 | |
| I0330 05:32:05.311642 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.5774 (* 0.3 = 0.77322 loss) | |
| I0330 05:32:05.311656 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.804794 (* 0.3 = 0.241438 loss) | |
| I0330 05:32:05.311668 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.490566 | |
| I0330 05:32:05.311681 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 05:32:05.311692 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.716981 | |
| I0330 05:32:05.311707 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.88339 (* 1 = 1.88339 loss) | |
| I0330 05:32:05.311730 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.590751 (* 1 = 0.590751 loss) | |
| I0330 05:32:05.311743 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:32:05.311755 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0367105 | |
| I0330 05:32:05.311769 10583 sgd_solver.cpp:106] Iteration 61500, lr = 0.01 | |
| I0330 05:34:14.494256 10583 solver.cpp:229] Iteration 62000, loss = 4.13494 | |
| I0330 05:34:14.494371 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.227273 | |
| I0330 05:34:14.494391 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 05:34:14.494405 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.454545 | |
| I0330 05:34:14.494421 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.3901 (* 0.3 = 0.717029 loss) | |
| I0330 05:34:14.494436 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.670985 (* 0.3 = 0.201295 loss) | |
| I0330 05:34:14.494448 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.272727 | |
| I0330 05:34:14.494462 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 05:34:14.494473 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.590909 | |
| I0330 05:34:14.494488 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12261 (* 0.3 = 0.636782 loss) | |
| I0330 05:34:14.494500 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.623743 (* 0.3 = 0.187123 loss) | |
| I0330 05:34:14.494513 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.568182 | |
| I0330 05:34:14.494524 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 05:34:14.494536 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.818182 | |
| I0330 05:34:14.494550 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.59095 (* 1 = 1.59095 loss) | |
| I0330 05:34:14.494565 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.452119 (* 1 = 0.452119 loss) | |
| I0330 05:34:14.494577 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:34:14.494588 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0661344 | |
| I0330 05:34:14.494601 10583 sgd_solver.cpp:106] Iteration 62000, lr = 0.01 | |
| I0330 05:36:23.901618 10583 solver.cpp:229] Iteration 62500, loss = 4.05571 | |
| I0330 05:36:23.901823 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0330 05:36:23.901844 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 05:36:23.901857 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.48 | |
| I0330 05:36:23.901882 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46907 (* 0.3 = 0.740721 loss) | |
| I0330 05:36:23.901897 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.724818 (* 0.3 = 0.217445 loss) | |
| I0330 05:36:23.901911 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.38 | |
| I0330 05:36:23.901923 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 05:36:23.901935 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0330 05:36:23.901949 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.28028 (* 0.3 = 0.684083 loss) | |
| I0330 05:36:23.901963 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.690082 (* 0.3 = 0.207024 loss) | |
| I0330 05:36:23.901976 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.62 | |
| I0330 05:36:23.901988 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 05:36:23.902000 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.74 | |
| I0330 05:36:23.902015 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.55486 (* 1 = 1.55486 loss) | |
| I0330 05:36:23.902029 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.469083 (* 1 = 0.469083 loss) | |
| I0330 05:36:23.902042 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:36:23.902053 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0803727 | |
| I0330 05:36:23.902067 10583 sgd_solver.cpp:106] Iteration 62500, lr = 0.01 | |
| I0330 05:38:33.141393 10583 solver.cpp:229] Iteration 63000, loss = 4.14752 | |
| I0330 05:38:33.141525 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535 | |
| I0330 05:38:33.141544 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 05:38:33.141558 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.302326 | |
| I0330 05:38:33.141576 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.7559 (* 0.3 = 0.826771 loss) | |
| I0330 05:38:33.141590 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.80943 (* 0.3 = 0.242829 loss) | |
| I0330 05:38:33.141603 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.348837 | |
| I0330 05:38:33.141615 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 05:38:33.141628 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.511628 | |
| I0330 05:38:33.141643 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.63171 (* 0.3 = 0.789514 loss) | |
| I0330 05:38:33.141656 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.7983 (* 0.3 = 0.23949 loss) | |
| I0330 05:38:33.141669 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.418605 | |
| I0330 05:38:33.141681 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0330 05:38:33.141696 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.651163 | |
| I0330 05:38:33.141711 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.15469 (* 1 = 2.15469 loss) | |
| I0330 05:38:33.141726 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.677518 (* 1 = 0.677518 loss) | |
| I0330 05:38:33.141737 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:38:33.141749 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0190625 | |
| I0330 05:38:33.141762 10583 sgd_solver.cpp:106] Iteration 63000, lr = 0.01 | |
| I0330 05:40:42.431376 10583 solver.cpp:229] Iteration 63500, loss = 4.02593 | |
| I0330 05:40:42.431548 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.163265 | |
| I0330 05:40:42.431568 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 05:40:42.431582 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.408163 | |
| I0330 05:40:42.431599 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.814 (* 0.3 = 0.844199 loss) | |
| I0330 05:40:42.431614 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.820014 (* 0.3 = 0.246004 loss) | |
| I0330 05:40:42.431627 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.306122 | |
| I0330 05:40:42.431648 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 05:40:42.431659 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.489796 | |
| I0330 05:40:42.431674 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34903 (* 0.3 = 0.704708 loss) | |
| I0330 05:40:42.431687 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.712132 (* 0.3 = 0.21364 loss) | |
| I0330 05:40:42.431699 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.489796 | |
| I0330 05:40:42.431711 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0330 05:40:42.431723 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.693878 | |
| I0330 05:40:42.431737 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.68388 (* 1 = 1.68388 loss) | |
| I0330 05:40:42.431751 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.553677 (* 1 = 0.553677 loss) | |
| I0330 05:40:42.431763 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:40:42.431776 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0566072 | |
| I0330 05:40:42.431797 10583 sgd_solver.cpp:106] Iteration 63500, lr = 0.01 | |
| I0330 05:42:51.793503 10583 solver.cpp:229] Iteration 64000, loss = 4.00964 | |
| I0330 05:42:51.793618 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.235294 | |
| I0330 05:42:51.793638 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 05:42:51.793651 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.45098 | |
| I0330 05:42:51.793668 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.6465 (* 0.3 = 0.79395 loss) | |
| I0330 05:42:51.793681 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.810927 (* 0.3 = 0.243278 loss) | |
| I0330 05:42:51.793694 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.27451 | |
| I0330 05:42:51.793706 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 05:42:51.793718 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.54902 | |
| I0330 05:42:51.793732 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.45438 (* 0.3 = 0.736314 loss) | |
| I0330 05:42:51.793747 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.775076 (* 0.3 = 0.232523 loss) | |
| I0330 05:42:51.793761 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.470588 | |
| I0330 05:42:51.793772 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 05:42:51.793784 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72549 | |
| I0330 05:42:51.793798 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.71921 (* 1 = 1.71921 loss) | |
| I0330 05:42:51.793812 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.528618 (* 1 = 0.528618 loss) | |
| I0330 05:42:51.793823 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:42:51.793835 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0226979 | |
| I0330 05:42:51.793848 10583 sgd_solver.cpp:106] Iteration 64000, lr = 0.01 | |
| I0330 05:45:01.055311 10583 solver.cpp:229] Iteration 64500, loss = 3.99586 | |
| I0330 05:45:01.055466 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.215686 | |
| I0330 05:45:01.055486 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 05:45:01.055500 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.490196 | |
| I0330 05:45:01.055517 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.60362 (* 0.3 = 0.781085 loss) | |
| I0330 05:45:01.055532 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.793079 (* 0.3 = 0.237924 loss) | |
| I0330 05:45:01.055544 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.235294 | |
| I0330 05:45:01.055557 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 05:45:01.055568 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.529412 | |
| I0330 05:45:01.055583 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.45389 (* 0.3 = 0.736167 loss) | |
| I0330 05:45:01.055598 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.745774 (* 0.3 = 0.223732 loss) | |
| I0330 05:45:01.055609 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.45098 | |
| I0330 05:45:01.055621 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 05:45:01.055634 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.627451 | |
| I0330 05:45:01.055647 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.96369 (* 1 = 1.96369 loss) | |
| I0330 05:45:01.055661 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.604899 (* 1 = 0.604899 loss) | |
| I0330 05:45:01.055675 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:45:01.055685 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0487788 | |
| I0330 05:45:01.055699 10583 sgd_solver.cpp:106] Iteration 64500, lr = 0.01 | |
| I0330 05:47:10.115026 10583 solver.cpp:338] Iteration 65000, Testing net (#0) | |
| I0330 05:47:39.894239 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 05:47:39.894289 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 05:47:39.894305 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759774 | |
| I0330 05:47:39.894318 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 05:47:39.894335 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:47:39.894350 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:47:39.894362 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 05:47:39.894374 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759774 | |
| I0330 05:47:39.894387 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 05:47:39.894399 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:47:39.894415 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 05:47:39.894428 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 05:47:39.894439 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759774 | |
| I0330 05:47:39.894451 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 05:47:39.894465 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:47:39.894480 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 05:47:39.894493 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 05:47:39.894505 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 05:47:40.044802 10583 solver.cpp:229] Iteration 65000, loss = 4.01768 | |
| I0330 05:47:40.044842 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.28 | |
| I0330 05:47:40.044858 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 05:47:40.044872 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.52 | |
| I0330 05:47:40.044886 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51236 (* 0.3 = 0.753708 loss) | |
| I0330 05:47:40.044901 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.766283 (* 0.3 = 0.229885 loss) | |
| I0330 05:47:40.044914 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.28 | |
| I0330 05:47:40.044925 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 05:47:40.044937 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.66 | |
| I0330 05:47:40.044951 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.29263 (* 0.3 = 0.687789 loss) | |
| I0330 05:47:40.044965 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.697677 (* 0.3 = 0.209303 loss) | |
| I0330 05:47:40.044978 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 05:47:40.044991 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 05:47:40.045002 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 05:47:40.045017 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.67665 (* 1 = 1.67665 loss) | |
| I0330 05:47:40.045030 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.500832 (* 1 = 0.500832 loss) | |
| I0330 05:47:40.045043 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:47:40.045054 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0454907 | |
| I0330 05:47:40.045068 10583 sgd_solver.cpp:106] Iteration 65000, lr = 0.01 | |
| I0330 05:49:49.109789 10583 solver.cpp:229] Iteration 65500, loss = 3.95281 | |
| I0330 05:49:49.109932 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.292683 | |
| I0330 05:49:49.109952 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 05:49:49.109966 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.463415 | |
| I0330 05:49:49.109982 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.38304 (* 0.3 = 0.714912 loss) | |
| I0330 05:49:49.109997 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.672113 (* 0.3 = 0.201634 loss) | |
| I0330 05:49:49.110013 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.463415 | |
| I0330 05:49:49.110024 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 05:49:49.110038 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.707317 | |
| I0330 05:49:49.110051 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.02551 (* 0.3 = 0.607653 loss) | |
| I0330 05:49:49.110065 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.587805 (* 0.3 = 0.176342 loss) | |
| I0330 05:49:49.110077 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.536585 | |
| I0330 05:49:49.110090 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 05:49:49.110101 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.780488 | |
| I0330 05:49:49.110116 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.4417 (* 1 = 1.4417 loss) | |
| I0330 05:49:49.110129 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.406163 (* 1 = 0.406163 loss) | |
| I0330 05:49:49.110141 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:49:49.110153 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0528646 | |
| I0330 05:49:49.110168 10583 sgd_solver.cpp:106] Iteration 65500, lr = 0.01 | |
| I0330 05:51:58.332412 10583 solver.cpp:229] Iteration 66000, loss = 3.89482 | |
| I0330 05:51:58.332520 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25641 | |
| I0330 05:51:58.332550 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 05:51:58.332562 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.538462 | |
| I0330 05:51:58.332579 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36896 (* 0.3 = 0.710688 loss) | |
| I0330 05:51:58.332594 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.627949 (* 0.3 = 0.188385 loss) | |
| I0330 05:51:58.332607 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.384615 | |
| I0330 05:51:58.332628 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 05:51:58.332641 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.641026 | |
| I0330 05:51:58.332656 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.16386 (* 0.3 = 0.649159 loss) | |
| I0330 05:51:58.332671 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.572048 (* 0.3 = 0.171614 loss) | |
| I0330 05:51:58.332684 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.615385 | |
| I0330 05:51:58.332695 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 05:51:58.332707 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.794872 | |
| I0330 05:51:58.332721 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.41501 (* 1 = 1.41501 loss) | |
| I0330 05:51:58.332736 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.393036 (* 1 = 0.393036 loss) | |
| I0330 05:51:58.332748 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:51:58.332761 10583 solver.cpp:245] Train net output #16: total_confidence = 0.13859 | |
| I0330 05:51:58.332772 10583 sgd_solver.cpp:106] Iteration 66000, lr = 0.01 | |
| I0330 05:54:07.616291 10583 solver.cpp:229] Iteration 66500, loss = 3.94299 | |
| I0330 05:54:07.616428 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.361702 | |
| I0330 05:54:07.616448 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 05:54:07.616462 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.531915 | |
| I0330 05:54:07.616478 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41362 (* 0.3 = 0.724086 loss) | |
| I0330 05:54:07.616493 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.74214 (* 0.3 = 0.222642 loss) | |
| I0330 05:54:07.616506 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.382979 | |
| I0330 05:54:07.616518 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 05:54:07.616529 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.659574 | |
| I0330 05:54:07.616544 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.02316 (* 0.3 = 0.606949 loss) | |
| I0330 05:54:07.616559 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.582905 (* 0.3 = 0.174872 loss) | |
| I0330 05:54:07.616570 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.553191 | |
| I0330 05:54:07.616582 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 05:54:07.616595 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0330 05:54:07.616610 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.25231 (* 1 = 1.25231 loss) | |
| I0330 05:54:07.616623 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.41583 (* 1 = 0.41583 loss) | |
| I0330 05:54:07.616636 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:54:07.616647 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0650765 | |
| I0330 05:54:07.616659 10583 sgd_solver.cpp:106] Iteration 66500, lr = 0.01 | |
| I0330 05:56:16.936574 10583 solver.cpp:229] Iteration 67000, loss = 3.89342 | |
| I0330 05:56:16.936741 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27907 | |
| I0330 05:56:16.936763 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 05:56:16.936782 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55814 | |
| I0330 05:56:16.936800 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.495 (* 0.3 = 0.7485 loss) | |
| I0330 05:56:16.936815 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.712522 (* 0.3 = 0.213757 loss) | |
| I0330 05:56:16.936826 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.27907 | |
| I0330 05:56:16.936839 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 05:56:16.936851 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.581395 | |
| I0330 05:56:16.936864 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.49155 (* 0.3 = 0.747465 loss) | |
| I0330 05:56:16.936879 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.699021 (* 0.3 = 0.209706 loss) | |
| I0330 05:56:16.936893 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.488372 | |
| I0330 05:56:16.936905 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 05:56:16.936916 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.744186 | |
| I0330 05:56:16.936930 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.5112 (* 1 = 1.5112 loss) | |
| I0330 05:56:16.936944 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.428525 (* 1 = 0.428525 loss) | |
| I0330 05:56:16.936957 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 05:56:16.936969 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0633659 | |
| I0330 05:56:16.936981 10583 sgd_solver.cpp:106] Iteration 67000, lr = 0.01 | |
| I0330 05:58:26.386880 10583 solver.cpp:229] Iteration 67500, loss = 3.90756 | |
| I0330 05:58:26.387060 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.234043 | |
| I0330 05:58:26.387081 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 05:58:26.387095 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489362 | |
| I0330 05:58:26.387112 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.49132 (* 0.3 = 0.747397 loss) | |
| I0330 05:58:26.387127 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.754446 (* 0.3 = 0.226334 loss) | |
| I0330 05:58:26.387140 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.340426 | |
| I0330 05:58:26.387152 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 05:58:26.387166 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.510638 | |
| I0330 05:58:26.387181 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.16387 (* 0.3 = 0.649161 loss) | |
| I0330 05:58:26.387195 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.70877 (* 0.3 = 0.212631 loss) | |
| I0330 05:58:26.387217 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.510638 | |
| I0330 05:58:26.387228 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 05:58:26.387240 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.702128 | |
| I0330 05:58:26.387255 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60014 (* 1 = 1.60014 loss) | |
| I0330 05:58:26.387269 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.528707 (* 1 = 0.528707 loss) | |
| I0330 05:58:26.387282 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 05:58:26.387295 10583 solver.cpp:245] Train net output #16: total_confidence = 0.129097 | |
| I0330 05:58:26.387307 10583 sgd_solver.cpp:106] Iteration 67500, lr = 0.01 | |
| I0330 06:00:35.458660 10583 solver.cpp:229] Iteration 68000, loss = 3.84289 | |
| I0330 06:00:35.458830 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395833 | |
| I0330 06:00:35.458853 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 06:00:35.458865 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5625 | |
| I0330 06:00:35.458890 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.47519 (* 0.3 = 0.742558 loss) | |
| I0330 06:00:35.458906 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.740169 (* 0.3 = 0.222051 loss) | |
| I0330 06:00:35.458919 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.354167 | |
| I0330 06:00:35.458931 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 06:00:35.458942 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 06:00:35.458956 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.16637 (* 0.3 = 0.649911 loss) | |
| I0330 06:00:35.458971 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.67935 (* 0.3 = 0.203805 loss) | |
| I0330 06:00:35.458997 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.541667 | |
| I0330 06:00:35.459010 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 06:00:35.459022 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.770833 | |
| I0330 06:00:35.459038 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.73727 (* 1 = 1.73727 loss) | |
| I0330 06:00:35.459051 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.54575 (* 1 = 0.54575 loss) | |
| I0330 06:00:35.459064 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:00:35.459076 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0581398 | |
| I0330 06:00:35.459089 10583 sgd_solver.cpp:106] Iteration 68000, lr = 0.01 | |
| I0330 06:02:44.786190 10583 solver.cpp:229] Iteration 68500, loss = 3.87836 | |
| I0330 06:02:44.786314 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.342105 | |
| I0330 06:02:44.786334 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 06:02:44.786356 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.552632 | |
| I0330 06:02:44.786372 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.31847 (* 0.3 = 0.695542 loss) | |
| I0330 06:02:44.786387 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.657136 (* 0.3 = 0.197141 loss) | |
| I0330 06:02:44.786401 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.447368 | |
| I0330 06:02:44.786418 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 06:02:44.786430 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.684211 | |
| I0330 06:02:44.786444 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.01649 (* 0.3 = 0.604948 loss) | |
| I0330 06:02:44.786458 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.558032 (* 0.3 = 0.16741 loss) | |
| I0330 06:02:44.786470 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.526316 | |
| I0330 06:02:44.786484 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 06:02:44.786495 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.736842 | |
| I0330 06:02:44.786509 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.36327 (* 1 = 1.36327 loss) | |
| I0330 06:02:44.786523 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.346075 (* 1 = 0.346075 loss) | |
| I0330 06:02:44.786535 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:02:44.786547 10583 solver.cpp:245] Train net output #16: total_confidence = 0.114198 | |
| I0330 06:02:44.786561 10583 sgd_solver.cpp:106] Iteration 68500, lr = 0.01 | |
| I0330 06:04:54.170752 10583 solver.cpp:229] Iteration 69000, loss = 3.86114 | |
| I0330 06:04:54.170878 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244444 | |
| I0330 06:04:54.170899 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 06:04:54.170912 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.555556 | |
| I0330 06:04:54.170928 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.35259 (* 0.3 = 0.705776 loss) | |
| I0330 06:04:54.170943 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.708941 (* 0.3 = 0.212682 loss) | |
| I0330 06:04:54.170956 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 06:04:54.170969 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 06:04:54.170980 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0330 06:04:54.170994 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.10067 (* 0.3 = 0.630201 loss) | |
| I0330 06:04:54.171021 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.674303 (* 0.3 = 0.202291 loss) | |
| I0330 06:04:54.171036 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.488889 | |
| I0330 06:04:54.171047 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 06:04:54.171059 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.733333 | |
| I0330 06:04:54.171073 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.62137 (* 1 = 1.62137 loss) | |
| I0330 06:04:54.171087 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.483934 (* 1 = 0.483934 loss) | |
| I0330 06:04:54.171100 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:04:54.171113 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0938777 | |
| I0330 06:04:54.171124 10583 sgd_solver.cpp:106] Iteration 69000, lr = 0.01 | |
| I0330 06:07:03.169770 10583 solver.cpp:229] Iteration 69500, loss = 3.83658 | |
| I0330 06:07:03.169922 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0330 06:07:03.169944 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 06:07:03.169957 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0330 06:07:03.169973 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48099 (* 0.3 = 0.744298 loss) | |
| I0330 06:07:03.169988 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.723324 (* 0.3 = 0.216997 loss) | |
| I0330 06:07:03.170001 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.244898 | |
| I0330 06:07:03.170013 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0330 06:07:03.170025 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.44898 | |
| I0330 06:07:03.170039 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.66485 (* 0.3 = 0.799454 loss) | |
| I0330 06:07:03.170053 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.790097 (* 0.3 = 0.237029 loss) | |
| I0330 06:07:03.170066 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.346939 | |
| I0330 06:07:03.170078 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0330 06:07:03.170090 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.693878 | |
| I0330 06:07:03.170104 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.04572 (* 1 = 2.04572 loss) | |
| I0330 06:07:03.170119 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.595889 (* 1 = 0.595889 loss) | |
| I0330 06:07:03.170131 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:07:03.170143 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0545347 | |
| I0330 06:07:03.170156 10583 sgd_solver.cpp:106] Iteration 69500, lr = 0.01 | |
| I0330 06:09:12.094079 10583 solver.cpp:338] Iteration 70000, Testing net (#0) | |
| I0330 06:09:41.872558 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 06:09:41.872611 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 06:09:41.872627 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759728 | |
| I0330 06:09:41.872640 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 06:09:41.872656 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:09:41.872673 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:09:41.872684 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 06:09:41.872696 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759728 | |
| I0330 06:09:41.872709 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 06:09:41.872722 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:09:41.872737 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:09:41.872750 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 06:09:41.872761 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759728 | |
| I0330 06:09:41.872773 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 06:09:41.872787 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:09:41.872802 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:09:41.872813 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 06:09:41.872825 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 06:09:42.025393 10583 solver.cpp:229] Iteration 70000, loss = 3.8842 | |
| I0330 06:09:42.025449 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.173077 | |
| I0330 06:09:42.025466 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 06:09:42.025480 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.442308 | |
| I0330 06:09:42.025496 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.79219 (* 0.3 = 0.837658 loss) | |
| I0330 06:09:42.025511 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.852764 (* 0.3 = 0.255829 loss) | |
| I0330 06:09:42.025523 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.288462 | |
| I0330 06:09:42.025537 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0330 06:09:42.025548 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.576923 | |
| I0330 06:09:42.025563 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.43276 (* 0.3 = 0.729829 loss) | |
| I0330 06:09:42.025578 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.752656 (* 0.3 = 0.225797 loss) | |
| I0330 06:09:42.025589 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.519231 | |
| I0330 06:09:42.025602 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 06:09:42.025614 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.711538 | |
| I0330 06:09:42.025629 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.6133 (* 1 = 1.6133 loss) | |
| I0330 06:09:42.025647 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.492048 (* 1 = 0.492048 loss) | |
| I0330 06:09:42.025672 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:09:42.025686 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0197102 | |
| I0330 06:09:42.025698 10583 sgd_solver.cpp:106] Iteration 70000, lr = 0.01 | |
| I0330 06:11:51.282305 10583 solver.cpp:229] Iteration 70500, loss = 3.84411 | |
| I0330 06:11:51.282444 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.361702 | |
| I0330 06:11:51.282464 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 06:11:51.282477 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.468085 | |
| I0330 06:11:51.282493 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.22339 (* 0.3 = 0.667016 loss) | |
| I0330 06:11:51.282508 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650057 (* 0.3 = 0.195017 loss) | |
| I0330 06:11:51.282521 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.382979 | |
| I0330 06:11:51.282533 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 06:11:51.282546 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.617021 | |
| I0330 06:11:51.282559 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.18462 (* 0.3 = 0.655385 loss) | |
| I0330 06:11:51.282574 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.637626 (* 0.3 = 0.191288 loss) | |
| I0330 06:11:51.282587 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.553191 | |
| I0330 06:11:51.282598 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 06:11:51.282611 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0330 06:11:51.282625 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.72813 (* 1 = 1.72813 loss) | |
| I0330 06:11:51.282639 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.489465 (* 1 = 0.489465 loss) | |
| I0330 06:11:51.282651 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:11:51.282663 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0817066 | |
| I0330 06:11:51.282675 10583 sgd_solver.cpp:106] Iteration 70500, lr = 0.01 | |
| I0330 06:14:00.484827 10583 solver.cpp:229] Iteration 71000, loss = 3.79513 | |
| I0330 06:14:00.484941 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.34 | |
| I0330 06:14:00.484961 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 06:14:00.484973 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0330 06:14:00.484992 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14783 (* 0.3 = 0.644348 loss) | |
| I0330 06:14:00.485007 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.659622 (* 0.3 = 0.197887 loss) | |
| I0330 06:14:00.485019 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 06:14:00.485031 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 06:14:00.485044 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0330 06:14:00.485059 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.74444 (* 0.3 = 0.523332 loss) | |
| I0330 06:14:00.485072 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.529568 (* 0.3 = 0.15887 loss) | |
| I0330 06:14:00.485085 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.74 | |
| I0330 06:14:00.485096 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 06:14:00.485108 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0330 06:14:00.485122 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.955816 (* 1 = 0.955816 loss) | |
| I0330 06:14:00.485137 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.305918 (* 1 = 0.305918 loss) | |
| I0330 06:14:00.485149 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:14:00.485164 10583 solver.cpp:245] Train net output #16: total_confidence = 0.150512 | |
| I0330 06:14:00.485177 10583 sgd_solver.cpp:106] Iteration 71000, lr = 0.01 | |
| I0330 06:16:09.571655 10583 solver.cpp:229] Iteration 71500, loss = 3.77912 | |
| I0330 06:16:09.571805 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.18 | |
| I0330 06:16:09.571825 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 06:16:09.571838 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.36 | |
| I0330 06:16:09.571856 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.53786 (* 0.3 = 0.761358 loss) | |
| I0330 06:16:09.571877 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.737295 (* 0.3 = 0.221189 loss) | |
| I0330 06:16:09.571889 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.36 | |
| I0330 06:16:09.571902 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 06:16:09.571913 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.66 | |
| I0330 06:16:09.571928 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04419 (* 0.3 = 0.613257 loss) | |
| I0330 06:16:09.571943 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.61584 (* 0.3 = 0.184752 loss) | |
| I0330 06:16:09.571954 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.46 | |
| I0330 06:16:09.571966 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 06:16:09.571979 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76 | |
| I0330 06:16:09.571992 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.51438 (* 1 = 1.51438 loss) | |
| I0330 06:16:09.572007 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.46403 (* 1 = 0.46403 loss) | |
| I0330 06:16:09.572019 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:16:09.572031 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0345757 | |
| I0330 06:16:09.572043 10583 sgd_solver.cpp:106] Iteration 71500, lr = 0.01 | |
| I0330 06:18:18.978373 10583 solver.cpp:229] Iteration 72000, loss = 3.82018 | |
| I0330 06:18:18.978499 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0330 06:18:18.978519 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 06:18:18.978533 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.422222 | |
| I0330 06:18:18.978549 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.55604 (* 0.3 = 0.766812 loss) | |
| I0330 06:18:18.978564 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.725653 (* 0.3 = 0.217696 loss) | |
| I0330 06:18:18.978575 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.422222 | |
| I0330 06:18:18.978588 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 06:18:18.978600 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0330 06:18:18.978613 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.08697 (* 0.3 = 0.62609 loss) | |
| I0330 06:18:18.978627 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.604594 (* 0.3 = 0.181378 loss) | |
| I0330 06:18:18.978641 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.711111 | |
| I0330 06:18:18.978652 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 06:18:18.978664 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.911111 | |
| I0330 06:18:18.978678 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.972882 (* 1 = 0.972882 loss) | |
| I0330 06:18:18.978693 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.277637 (* 1 = 0.277637 loss) | |
| I0330 06:18:18.978704 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 06:18:18.978718 10583 solver.cpp:245] Train net output #16: total_confidence = 0.108835 | |
| I0330 06:18:18.978729 10583 sgd_solver.cpp:106] Iteration 72000, lr = 0.01 | |
| I0330 06:20:28.212244 10583 solver.cpp:229] Iteration 72500, loss = 3.74273 | |
| I0330 06:20:28.212389 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326087 | |
| I0330 06:20:28.212410 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 06:20:28.212424 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.586957 | |
| I0330 06:20:28.212450 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.47145 (* 0.3 = 0.741436 loss) | |
| I0330 06:20:28.212463 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.726918 (* 0.3 = 0.218075 loss) | |
| I0330 06:20:28.212476 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.434783 | |
| I0330 06:20:28.212488 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 06:20:28.212501 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.586957 | |
| I0330 06:20:28.212514 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12246 (* 0.3 = 0.636737 loss) | |
| I0330 06:20:28.212529 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.652372 (* 0.3 = 0.195712 loss) | |
| I0330 06:20:28.212541 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.608696 | |
| I0330 06:20:28.212553 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 06:20:28.212565 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76087 | |
| I0330 06:20:28.212579 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.52096 (* 1 = 1.52096 loss) | |
| I0330 06:20:28.212594 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.487648 (* 1 = 0.487648 loss) | |
| I0330 06:20:28.212606 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:20:28.212617 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0895323 | |
| I0330 06:20:28.212630 10583 sgd_solver.cpp:106] Iteration 72500, lr = 0.01 | |
| I0330 06:22:37.166324 10583 solver.cpp:229] Iteration 73000, loss = 3.75431 | |
| I0330 06:22:37.166443 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12963 | |
| I0330 06:22:37.166465 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0330 06:22:37.166477 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.351852 | |
| I0330 06:22:37.166493 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.10575 (* 0.3 = 0.931726 loss) | |
| I0330 06:22:37.166508 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.991052 (* 0.3 = 0.297316 loss) | |
| I0330 06:22:37.166522 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.259259 | |
| I0330 06:22:37.166533 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 06:22:37.166545 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.388889 | |
| I0330 06:22:37.166559 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.10999 (* 0.3 = 0.932996 loss) | |
| I0330 06:22:37.166574 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.99126 (* 0.3 = 0.297378 loss) | |
| I0330 06:22:37.166585 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.333333 | |
| I0330 06:22:37.166597 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0330 06:22:37.166610 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.537037 | |
| I0330 06:22:37.166623 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.37653 (* 1 = 2.37653 loss) | |
| I0330 06:22:37.166638 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.796893 (* 1 = 0.796893 loss) | |
| I0330 06:22:37.166651 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:22:37.166662 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0703573 | |
| I0330 06:22:37.166674 10583 sgd_solver.cpp:106] Iteration 73000, lr = 0.01 | |
| I0330 06:24:46.347617 10583 solver.cpp:229] Iteration 73500, loss = 3.74978 | |
| I0330 06:24:46.347754 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.347826 | |
| I0330 06:24:46.347774 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 06:24:46.347787 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.565217 | |
| I0330 06:24:46.347812 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5416 (* 0.3 = 0.762481 loss) | |
| I0330 06:24:46.347836 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.713072 (* 0.3 = 0.213922 loss) | |
| I0330 06:24:46.347851 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.347826 | |
| I0330 06:24:46.347862 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 06:24:46.347874 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.695652 | |
| I0330 06:24:46.347888 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12905 (* 0.3 = 0.638714 loss) | |
| I0330 06:24:46.347903 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.616398 (* 0.3 = 0.184919 loss) | |
| I0330 06:24:46.347916 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673913 | |
| I0330 06:24:46.347929 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 06:24:46.347940 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0330 06:24:46.347954 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.22756 (* 1 = 1.22756 loss) | |
| I0330 06:24:46.347968 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.366182 (* 1 = 0.366182 loss) | |
| I0330 06:24:46.347980 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:24:46.347992 10583 solver.cpp:245] Train net output #16: total_confidence = 0.159705 | |
| I0330 06:24:46.348006 10583 sgd_solver.cpp:106] Iteration 73500, lr = 0.01 | |
| I0330 06:26:55.418241 10583 solver.cpp:229] Iteration 74000, loss = 3.70441 | |
| I0330 06:26:55.418376 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 06:26:55.418398 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 06:26:55.418411 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.4375 | |
| I0330 06:26:55.418428 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4421 (* 0.3 = 0.732631 loss) | |
| I0330 06:26:55.418442 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.757676 (* 0.3 = 0.227303 loss) | |
| I0330 06:26:55.418455 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0330 06:26:55.418468 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 06:26:55.418480 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.541667 | |
| I0330 06:26:55.418494 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.09007 (* 0.3 = 0.627022 loss) | |
| I0330 06:26:55.418509 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.660822 (* 0.3 = 0.198247 loss) | |
| I0330 06:26:55.418520 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.583333 | |
| I0330 06:26:55.418532 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 06:26:55.418545 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0330 06:26:55.418558 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.39976 (* 1 = 1.39976 loss) | |
| I0330 06:26:55.418572 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.41589 (* 1 = 0.41589 loss) | |
| I0330 06:26:55.418584 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:26:55.418596 10583 solver.cpp:245] Train net output #16: total_confidence = 0.145525 | |
| I0330 06:26:55.418608 10583 sgd_solver.cpp:106] Iteration 74000, lr = 0.01 | |
| I0330 06:29:04.535599 10583 solver.cpp:229] Iteration 74500, loss = 3.70546 | |
| I0330 06:29:04.535744 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.137255 | |
| I0330 06:29:04.535763 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 06:29:04.535778 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.588235 | |
| I0330 06:29:04.535795 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36131 (* 0.3 = 0.708394 loss) | |
| I0330 06:29:04.535815 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.72237 (* 0.3 = 0.216711 loss) | |
| I0330 06:29:04.535827 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.352941 | |
| I0330 06:29:04.535840 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 06:29:04.535852 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.705882 | |
| I0330 06:29:04.535866 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.02381 (* 0.3 = 0.607142 loss) | |
| I0330 06:29:04.535881 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.641216 (* 0.3 = 0.192365 loss) | |
| I0330 06:29:04.535892 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.54902 | |
| I0330 06:29:04.535904 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 06:29:04.535917 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.843137 | |
| I0330 06:29:04.535931 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.30713 (* 1 = 1.30713 loss) | |
| I0330 06:29:04.535945 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.420688 (* 1 = 0.420688 loss) | |
| I0330 06:29:04.535958 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:29:04.535969 10583 solver.cpp:245] Train net output #16: total_confidence = 0.132997 | |
| I0330 06:29:04.535981 10583 sgd_solver.cpp:106] Iteration 74500, lr = 0.01 | |
| I0330 06:31:13.726052 10583 solver.cpp:338] Iteration 75000, Testing net (#0) | |
| I0330 06:31:43.785411 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 06:31:43.785537 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 06:31:43.785555 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759592 | |
| I0330 06:31:43.785568 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 06:31:43.785585 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:31:43.785601 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:31:43.785614 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 06:31:43.785625 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759592 | |
| I0330 06:31:43.785637 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 06:31:43.785652 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:31:43.785666 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:31:43.785678 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 06:31:43.785691 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759592 | |
| I0330 06:31:43.785701 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 06:31:43.785715 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:31:43.785729 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:31:43.785742 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 06:31:43.785754 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 06:31:43.938215 10583 solver.cpp:229] Iteration 75000, loss = 3.75998 | |
| I0330 06:31:43.938293 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340909 | |
| I0330 06:31:43.938311 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 06:31:43.938325 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0330 06:31:43.938343 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.1991 (* 0.3 = 0.659731 loss) | |
| I0330 06:31:43.938357 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.706509 (* 0.3 = 0.211953 loss) | |
| I0330 06:31:43.938370 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431818 | |
| I0330 06:31:43.938382 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 06:31:43.938395 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.727273 | |
| I0330 06:31:43.938408 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.81827 (* 0.3 = 0.545482 loss) | |
| I0330 06:31:43.938422 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.632597 (* 0.3 = 0.189779 loss) | |
| I0330 06:31:43.938436 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.590909 | |
| I0330 06:31:43.938447 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 06:31:43.938460 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0330 06:31:43.938475 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3818 (* 1 = 1.3818 loss) | |
| I0330 06:31:43.938489 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.553108 (* 1 = 0.553108 loss) | |
| I0330 06:31:43.938503 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:31:43.938514 10583 solver.cpp:245] Train net output #16: total_confidence = 0.192539 | |
| I0330 06:31:43.938527 10583 sgd_solver.cpp:106] Iteration 75000, lr = 0.01 | |
| I0330 06:33:53.549507 10583 solver.cpp:229] Iteration 75500, loss = 3.62702 | |
| I0330 06:33:53.549661 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.170213 | |
| I0330 06:33:53.549682 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 06:33:53.549697 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489362 | |
| I0330 06:33:53.549713 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58233 (* 0.3 = 0.774698 loss) | |
| I0330 06:33:53.549728 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.793021 (* 0.3 = 0.237906 loss) | |
| I0330 06:33:53.549741 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.340426 | |
| I0330 06:33:53.549753 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 06:33:53.549765 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.574468 | |
| I0330 06:33:53.549779 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.41965 (* 0.3 = 0.725896 loss) | |
| I0330 06:33:53.549793 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.73188 (* 0.3 = 0.219564 loss) | |
| I0330 06:33:53.549805 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.531915 | |
| I0330 06:33:53.549818 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 06:33:53.549829 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.765957 | |
| I0330 06:33:53.549844 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.53766 (* 1 = 1.53766 loss) | |
| I0330 06:33:53.549859 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.503371 (* 1 = 0.503371 loss) | |
| I0330 06:33:53.549870 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:33:53.549883 10583 solver.cpp:245] Train net output #16: total_confidence = 0.1193 | |
| I0330 06:33:53.549896 10583 sgd_solver.cpp:106] Iteration 75500, lr = 0.01 | |
| I0330 06:36:03.054764 10583 solver.cpp:229] Iteration 76000, loss = 3.64254 | |
| I0330 06:36:03.054906 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0330 06:36:03.054927 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 06:36:03.054940 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0330 06:36:03.054965 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.33215 (* 0.3 = 0.699645 loss) | |
| I0330 06:36:03.054978 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.682641 (* 0.3 = 0.204792 loss) | |
| I0330 06:36:03.054991 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.387755 | |
| I0330 06:36:03.055003 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 06:36:03.055016 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0330 06:36:03.055045 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88944 (* 0.3 = 0.566832 loss) | |
| I0330 06:36:03.055061 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.565762 (* 0.3 = 0.169728 loss) | |
| I0330 06:36:03.055074 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.734694 | |
| I0330 06:36:03.055086 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 06:36:03.055099 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.897959 | |
| I0330 06:36:03.055114 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.862242 (* 1 = 0.862242 loss) | |
| I0330 06:36:03.055127 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.258052 (* 1 = 0.258052 loss) | |
| I0330 06:36:03.055140 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 06:36:03.055151 10583 solver.cpp:245] Train net output #16: total_confidence = 0.205191 | |
| I0330 06:36:03.055166 10583 sgd_solver.cpp:106] Iteration 76000, lr = 0.01 | |
| I0330 06:38:12.150188 10583 solver.cpp:229] Iteration 76500, loss = 3.68009 | |
| I0330 06:38:12.150302 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.196078 | |
| I0330 06:38:12.150322 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 06:38:12.150336 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.392157 | |
| I0330 06:38:12.150352 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65016 (* 0.3 = 0.795049 loss) | |
| I0330 06:38:12.150367 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.802753 (* 0.3 = 0.240826 loss) | |
| I0330 06:38:12.150379 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.352941 | |
| I0330 06:38:12.150391 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 06:38:12.150404 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.588235 | |
| I0330 06:38:12.150418 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.22671 (* 0.3 = 0.668013 loss) | |
| I0330 06:38:12.150432 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.678606 (* 0.3 = 0.203582 loss) | |
| I0330 06:38:12.150444 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0330 06:38:12.150456 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 06:38:12.150467 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.803922 | |
| I0330 06:38:12.150482 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.37023 (* 1 = 1.37023 loss) | |
| I0330 06:38:12.150496 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.411758 (* 1 = 0.411758 loss) | |
| I0330 06:38:12.150508 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 06:38:12.150521 10583 solver.cpp:245] Train net output #16: total_confidence = 0.160784 | |
| I0330 06:38:12.150532 10583 sgd_solver.cpp:106] Iteration 76500, lr = 0.01 | |
| I0330 06:40:21.245780 10583 solver.cpp:229] Iteration 77000, loss = 3.662 | |
| I0330 06:40:21.245925 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.304348 | |
| I0330 06:40:21.245946 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 06:40:21.245959 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0330 06:40:21.245983 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46816 (* 0.3 = 0.740447 loss) | |
| I0330 06:40:21.245998 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.786591 (* 0.3 = 0.235977 loss) | |
| I0330 06:40:21.246011 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.391304 | |
| I0330 06:40:21.246023 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 06:40:21.246036 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673913 | |
| I0330 06:40:21.246049 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95456 (* 0.3 = 0.58637 loss) | |
| I0330 06:40:21.246063 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.637586 (* 0.3 = 0.191276 loss) | |
| I0330 06:40:21.246075 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.543478 | |
| I0330 06:40:21.246088 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 06:40:21.246099 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826087 | |
| I0330 06:40:21.246114 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26066 (* 1 = 1.26066 loss) | |
| I0330 06:40:21.246129 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.432901 (* 1 = 0.432901 loss) | |
| I0330 06:40:21.246141 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 06:40:21.246153 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0766124 | |
| I0330 06:40:21.246168 10583 sgd_solver.cpp:106] Iteration 77000, lr = 0.01 | |
| I0330 06:42:30.333742 10583 solver.cpp:229] Iteration 77500, loss = 3.66937 | |
| I0330 06:42:30.333875 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.346939 | |
| I0330 06:42:30.333895 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 06:42:30.333909 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0330 06:42:30.333925 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23027 (* 0.3 = 0.66908 loss) | |
| I0330 06:42:30.333943 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.658439 (* 0.3 = 0.197532 loss) | |
| I0330 06:42:30.333956 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.367347 | |
| I0330 06:42:30.333968 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 06:42:30.333981 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.612245 | |
| I0330 06:42:30.333995 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06567 (* 0.3 = 0.619702 loss) | |
| I0330 06:42:30.334009 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.619843 (* 0.3 = 0.185953 loss) | |
| I0330 06:42:30.334022 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.55102 | |
| I0330 06:42:30.334034 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 06:42:30.334045 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.836735 | |
| I0330 06:42:30.334060 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.37297 (* 1 = 1.37297 loss) | |
| I0330 06:42:30.334074 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.402229 (* 1 = 0.402229 loss) | |
| I0330 06:42:30.334086 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:42:30.334098 10583 solver.cpp:245] Train net output #16: total_confidence = 0.188371 | |
| I0330 06:42:30.334111 10583 sgd_solver.cpp:106] Iteration 77500, lr = 0.01 | |
| I0330 06:44:39.755426 10583 solver.cpp:229] Iteration 78000, loss = 3.60447 | |
| I0330 06:44:39.755594 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340909 | |
| I0330 06:44:39.755616 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 06:44:39.755631 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.636364 | |
| I0330 06:44:39.755647 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10458 (* 0.3 = 0.631373 loss) | |
| I0330 06:44:39.755662 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.6615 (* 0.3 = 0.19845 loss) | |
| I0330 06:44:39.755676 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431818 | |
| I0330 06:44:39.755688 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 06:44:39.755708 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.704545 | |
| I0330 06:44:39.755723 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83138 (* 0.3 = 0.549413 loss) | |
| I0330 06:44:39.755738 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.587845 (* 0.3 = 0.176353 loss) | |
| I0330 06:44:39.755749 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0330 06:44:39.755761 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 06:44:39.755774 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.840909 | |
| I0330 06:44:39.755787 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.23706 (* 1 = 1.23706 loss) | |
| I0330 06:44:39.755801 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.378058 (* 1 = 0.378058 loss) | |
| I0330 06:44:39.755813 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:44:39.755826 10583 solver.cpp:245] Train net output #16: total_confidence = 0.134176 | |
| I0330 06:44:39.755843 10583 sgd_solver.cpp:106] Iteration 78000, lr = 0.01 | |
| I0330 06:46:49.103253 10583 solver.cpp:229] Iteration 78500, loss = 3.64958 | |
| I0330 06:46:49.103360 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.304348 | |
| I0330 06:46:49.103381 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 06:46:49.103394 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.586957 | |
| I0330 06:46:49.103412 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48685 (* 0.3 = 0.746056 loss) | |
| I0330 06:46:49.103427 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.693389 (* 0.3 = 0.208017 loss) | |
| I0330 06:46:49.103440 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.521739 | |
| I0330 06:46:49.103452 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 06:46:49.103464 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.717391 | |
| I0330 06:46:49.103478 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96984 (* 0.3 = 0.590953 loss) | |
| I0330 06:46:49.103492 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.559239 (* 0.3 = 0.167772 loss) | |
| I0330 06:46:49.103505 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.695652 | |
| I0330 06:46:49.103518 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 06:46:49.103529 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804348 | |
| I0330 06:46:49.103543 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.32869 (* 1 = 1.32869 loss) | |
| I0330 06:46:49.103557 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.385125 (* 1 = 0.385125 loss) | |
| I0330 06:46:49.103569 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:46:49.103582 10583 solver.cpp:245] Train net output #16: total_confidence = 0.146448 | |
| I0330 06:46:49.103595 10583 sgd_solver.cpp:106] Iteration 78500, lr = 0.01 | |
| I0330 06:48:58.687623 10583 solver.cpp:229] Iteration 79000, loss = 3.60196 | |
| I0330 06:48:58.687770 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372093 | |
| I0330 06:48:58.687790 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 06:48:58.687803 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55814 | |
| I0330 06:48:58.687822 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2223 (* 0.3 = 0.666691 loss) | |
| I0330 06:48:58.687852 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.659881 (* 0.3 = 0.197964 loss) | |
| I0330 06:48:58.687866 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.534884 | |
| I0330 06:48:58.687878 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 06:48:58.687891 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.767442 | |
| I0330 06:48:58.687904 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66166 (* 0.3 = 0.498498 loss) | |
| I0330 06:48:58.687918 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.519633 (* 0.3 = 0.15589 loss) | |
| I0330 06:48:58.687930 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.697674 | |
| I0330 06:48:58.687942 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 06:48:58.687954 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.906977 | |
| I0330 06:48:58.687970 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.988752 (* 1 = 0.988752 loss) | |
| I0330 06:48:58.687984 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.277948 (* 1 = 0.277948 loss) | |
| I0330 06:48:58.687996 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 06:48:58.688009 10583 solver.cpp:245] Train net output #16: total_confidence = 0.186897 | |
| I0330 06:48:58.688020 10583 sgd_solver.cpp:106] Iteration 79000, lr = 0.01 | |
| I0330 06:51:07.930310 10583 solver.cpp:229] Iteration 79500, loss = 3.66947 | |
| I0330 06:51:07.930510 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.295455 | |
| I0330 06:51:07.930532 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 06:51:07.930546 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.590909 | |
| I0330 06:51:07.930563 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.22654 (* 0.3 = 0.667961 loss) | |
| I0330 06:51:07.930578 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650834 (* 0.3 = 0.19525 loss) | |
| I0330 06:51:07.930591 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431818 | |
| I0330 06:51:07.930604 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 06:51:07.930616 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.681818 | |
| I0330 06:51:07.930631 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9938 (* 0.3 = 0.59814 loss) | |
| I0330 06:51:07.930645 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.627527 (* 0.3 = 0.188258 loss) | |
| I0330 06:51:07.930658 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0330 06:51:07.930670 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 06:51:07.930682 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.772727 | |
| I0330 06:51:07.930696 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26952 (* 1 = 1.26952 loss) | |
| I0330 06:51:07.930711 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.376419 (* 1 = 0.376419 loss) | |
| I0330 06:51:07.930723 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:51:07.930735 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0975973 | |
| I0330 06:51:07.930748 10583 sgd_solver.cpp:106] Iteration 79500, lr = 0.01 | |
| I0330 06:53:17.584300 10583 solver.cpp:338] Iteration 80000, Testing net (#0) | |
| I0330 06:53:47.361105 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 06:53:47.361155 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 06:53:47.361174 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.76041 | |
| I0330 06:53:47.361188 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 06:53:47.361204 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:53:47.361220 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:53:47.361233 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 06:53:47.361245 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.76041 | |
| I0330 06:53:47.361258 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 06:53:47.361271 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:53:47.361287 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 06:53:47.361299 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 06:53:47.361311 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.76041 | |
| I0330 06:53:47.361323 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 06:53:47.361337 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:53:47.361351 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 06:53:47.361363 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 06:53:47.361376 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 06:53:47.512531 10583 solver.cpp:229] Iteration 80000, loss = 3.60989 | |
| I0330 06:53:47.512569 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.210526 | |
| I0330 06:53:47.512585 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 06:53:47.512598 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.421053 | |
| I0330 06:53:47.512614 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.66656 (* 0.3 = 0.799967 loss) | |
| I0330 06:53:47.512627 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.791672 (* 0.3 = 0.237502 loss) | |
| I0330 06:53:47.512642 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.342105 | |
| I0330 06:53:47.512655 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 06:53:47.512667 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.605263 | |
| I0330 06:53:47.512681 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.14639 (* 0.3 = 0.643918 loss) | |
| I0330 06:53:47.512696 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.621496 (* 0.3 = 0.186449 loss) | |
| I0330 06:53:47.512708 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.631579 | |
| I0330 06:53:47.512720 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 06:53:47.512732 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.815789 | |
| I0330 06:53:47.512747 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29653 (* 1 = 1.29653 loss) | |
| I0330 06:53:47.512760 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.356984 (* 1 = 0.356984 loss) | |
| I0330 06:53:47.512773 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 06:53:47.512784 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0909985 | |
| I0330 06:53:47.512796 10583 sgd_solver.cpp:106] Iteration 80000, lr = 0.01 | |
| I0330 06:55:56.663338 10583 solver.cpp:229] Iteration 80500, loss = 3.57637 | |
| I0330 06:55:56.663513 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.390244 | |
| I0330 06:55:56.663542 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 06:55:56.663573 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.682927 | |
| I0330 06:55:56.663600 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93278 (* 0.3 = 0.579835 loss) | |
| I0330 06:55:56.663625 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.512471 (* 0.3 = 0.153741 loss) | |
| I0330 06:55:56.663646 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.365854 | |
| I0330 06:55:56.663668 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 06:55:56.663691 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.756098 | |
| I0330 06:55:56.663714 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66155 (* 0.3 = 0.498464 loss) | |
| I0330 06:55:56.663739 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.446728 (* 0.3 = 0.134018 loss) | |
| I0330 06:55:56.663760 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.731707 | |
| I0330 06:55:56.663781 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 06:55:56.663801 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.878049 | |
| I0330 06:55:56.663826 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.91886 (* 1 = 0.91886 loss) | |
| I0330 06:55:56.663852 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.250507 (* 1 = 0.250507 loss) | |
| I0330 06:55:56.663873 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:55:56.663894 10583 solver.cpp:245] Train net output #16: total_confidence = 0.142923 | |
| I0330 06:55:56.663915 10583 sgd_solver.cpp:106] Iteration 80500, lr = 0.01 | |
| I0330 06:58:06.019574 10583 solver.cpp:229] Iteration 81000, loss = 3.5876 | |
| I0330 06:58:06.019690 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0330 06:58:06.019708 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 06:58:06.019721 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.479167 | |
| I0330 06:58:06.019738 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39157 (* 0.3 = 0.71747 loss) | |
| I0330 06:58:06.019753 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.712565 (* 0.3 = 0.213769 loss) | |
| I0330 06:58:06.019765 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 06:58:06.019778 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 06:58:06.019790 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.708333 | |
| I0330 06:58:06.019804 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95486 (* 0.3 = 0.586458 loss) | |
| I0330 06:58:06.019819 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.568673 (* 0.3 = 0.170602 loss) | |
| I0330 06:58:06.019831 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.645833 | |
| I0330 06:58:06.019843 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 06:58:06.019855 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.854167 | |
| I0330 06:58:06.019870 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.02584 (* 1 = 1.02584 loss) | |
| I0330 06:58:06.019883 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.302735 (* 1 = 0.302735 loss) | |
| I0330 06:58:06.019896 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 06:58:06.019907 10583 solver.cpp:245] Train net output #16: total_confidence = 0.107604 | |
| I0330 06:58:06.019920 10583 sgd_solver.cpp:106] Iteration 81000, lr = 0.01 | |
| I0330 07:00:15.577314 10583 solver.cpp:229] Iteration 81500, loss = 3.58278 | |
| I0330 07:00:15.577463 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 07:00:15.577493 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 07:00:15.577517 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0330 07:00:15.577549 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42985 (* 0.3 = 0.728955 loss) | |
| I0330 07:00:15.577575 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.667401 (* 0.3 = 0.20022 loss) | |
| I0330 07:00:15.577597 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404762 | |
| I0330 07:00:15.577620 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 07:00:15.577642 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 07:00:15.577667 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06801 (* 0.3 = 0.620403 loss) | |
| I0330 07:00:15.577692 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.610802 (* 0.3 = 0.18324 loss) | |
| I0330 07:00:15.577713 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.595238 | |
| I0330 07:00:15.577735 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 07:00:15.577757 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.714286 | |
| I0330 07:00:15.577783 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.50703 (* 1 = 1.50703 loss) | |
| I0330 07:00:15.577808 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.419082 (* 1 = 0.419082 loss) | |
| I0330 07:00:15.577831 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:00:15.577852 10583 solver.cpp:245] Train net output #16: total_confidence = 0.156101 | |
| I0330 07:00:15.577873 10583 sgd_solver.cpp:106] Iteration 81500, lr = 0.01 | |
| I0330 07:02:24.921845 10583 solver.cpp:229] Iteration 82000, loss = 3.55325 | |
| I0330 07:02:24.921968 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.428571 | |
| I0330 07:02:24.921990 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 07:02:24.922003 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 07:02:24.922019 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.55458 (* 0.3 = 0.766374 loss) | |
| I0330 07:02:24.922034 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.707602 (* 0.3 = 0.212281 loss) | |
| I0330 07:02:24.922047 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.47619 | |
| I0330 07:02:24.922060 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 07:02:24.922072 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.690476 | |
| I0330 07:02:24.922086 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.99073 (* 0.3 = 0.597219 loss) | |
| I0330 07:02:24.922101 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.5248 (* 0.3 = 0.15744 loss) | |
| I0330 07:02:24.922113 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0330 07:02:24.922125 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 07:02:24.922137 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97619 | |
| I0330 07:02:24.922152 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.84531 (* 1 = 0.84531 loss) | |
| I0330 07:02:24.922169 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.252929 (* 1 = 0.252929 loss) | |
| I0330 07:02:24.922183 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:02:24.922195 10583 solver.cpp:245] Train net output #16: total_confidence = 0.188148 | |
| I0330 07:02:24.922207 10583 sgd_solver.cpp:106] Iteration 82000, lr = 0.01 | |
| I0330 07:04:33.845580 10583 solver.cpp:229] Iteration 82500, loss = 3.62173 | |
| I0330 07:04:33.845747 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.302326 | |
| I0330 07:04:33.845777 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 07:04:33.845810 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.581395 | |
| I0330 07:04:33.845839 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42313 (* 0.3 = 0.726938 loss) | |
| I0330 07:04:33.845866 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.672237 (* 0.3 = 0.201671 loss) | |
| I0330 07:04:33.845888 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.348837 | |
| I0330 07:04:33.845911 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 07:04:33.845933 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.55814 | |
| I0330 07:04:33.845958 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.31918 (* 0.3 = 0.695754 loss) | |
| I0330 07:04:33.845984 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.631406 (* 0.3 = 0.189422 loss) | |
| I0330 07:04:33.846005 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.465116 | |
| I0330 07:04:33.846026 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 07:04:33.846046 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.697674 | |
| I0330 07:04:33.846073 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82787 (* 1 = 1.82787 loss) | |
| I0330 07:04:33.846098 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.506657 (* 1 = 0.506657 loss) | |
| I0330 07:04:33.846120 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 07:04:33.846140 10583 solver.cpp:245] Train net output #16: total_confidence = 0.102561 | |
| I0330 07:04:33.846165 10583 sgd_solver.cpp:106] Iteration 82500, lr = 0.01 | |
| I0330 07:06:43.212390 10583 solver.cpp:229] Iteration 83000, loss = 3.53607 | |
| I0330 07:06:43.212539 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.241379 | |
| I0330 07:06:43.212561 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 07:06:43.212574 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.396552 | |
| I0330 07:06:43.212591 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.77784 (* 0.3 = 0.833351 loss) | |
| I0330 07:06:43.212606 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.938054 (* 0.3 = 0.281416 loss) | |
| I0330 07:06:43.212620 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.293103 | |
| I0330 07:06:43.212631 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 07:06:43.212643 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.465517 | |
| I0330 07:06:43.212657 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.53416 (* 0.3 = 0.760249 loss) | |
| I0330 07:06:43.212671 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.841734 (* 0.3 = 0.25252 loss) | |
| I0330 07:06:43.212683 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.482759 | |
| I0330 07:06:43.212697 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0330 07:06:43.212708 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.689655 | |
| I0330 07:06:43.212723 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.74493 (* 1 = 1.74493 loss) | |
| I0330 07:06:43.212736 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.593832 (* 1 = 0.593832 loss) | |
| I0330 07:06:43.212749 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:06:43.212760 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0808701 | |
| I0330 07:06:43.212774 10583 sgd_solver.cpp:106] Iteration 83000, lr = 0.01 | |
| I0330 07:08:52.624068 10583 solver.cpp:229] Iteration 83500, loss = 3.53858 | |
| I0330 07:08:52.624224 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.294118 | |
| I0330 07:08:52.624254 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 07:08:52.624277 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509804 | |
| I0330 07:08:52.624307 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.7041 (* 0.3 = 0.81123 loss) | |
| I0330 07:08:52.624332 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.809756 (* 0.3 = 0.242927 loss) | |
| I0330 07:08:52.624351 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.392157 | |
| I0330 07:08:52.624372 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 07:08:52.624393 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.588235 | |
| I0330 07:08:52.624428 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.49516 (* 0.3 = 0.748549 loss) | |
| I0330 07:08:52.624455 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.750432 (* 0.3 = 0.22513 loss) | |
| I0330 07:08:52.624485 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0330 07:08:52.624507 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 07:08:52.624527 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.803922 | |
| I0330 07:08:52.624559 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.50686 (* 1 = 1.50686 loss) | |
| I0330 07:08:52.624584 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.456305 (* 1 = 0.456305 loss) | |
| I0330 07:08:52.624605 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 07:08:52.624626 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0945212 | |
| I0330 07:08:52.624649 10583 sgd_solver.cpp:106] Iteration 83500, lr = 0.01 | |
| I0330 07:11:02.018640 10583 solver.cpp:229] Iteration 84000, loss = 3.49847 | |
| I0330 07:11:02.018757 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.24 | |
| I0330 07:11:02.018795 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 07:11:02.018817 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0330 07:11:02.018846 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.29871 (* 0.3 = 0.689613 loss) | |
| I0330 07:11:02.018872 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.679206 (* 0.3 = 0.203762 loss) | |
| I0330 07:11:02.018901 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.46 | |
| I0330 07:11:02.018925 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 07:11:02.018947 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.7 | |
| I0330 07:11:02.018986 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03683 (* 0.3 = 0.61105 loss) | |
| I0330 07:11:02.019017 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.602793 (* 0.3 = 0.180838 loss) | |
| I0330 07:11:02.019038 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6 | |
| I0330 07:11:02.019060 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 07:11:02.019083 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0330 07:11:02.019109 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.11702 (* 1 = 1.11702 loss) | |
| I0330 07:11:02.019134 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.322825 (* 1 = 0.322825 loss) | |
| I0330 07:11:02.019155 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:11:02.019181 10583 solver.cpp:245] Train net output #16: total_confidence = 0.146313 | |
| I0330 07:11:02.019202 10583 sgd_solver.cpp:106] Iteration 84000, lr = 0.01 | |
| I0330 07:13:11.441602 10583 solver.cpp:229] Iteration 84500, loss = 3.54637 | |
| I0330 07:13:11.441750 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.276596 | |
| I0330 07:13:11.441769 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 07:13:11.441783 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.574468 | |
| I0330 07:13:11.441799 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.56667 (* 0.3 = 0.77 loss) | |
| I0330 07:13:11.441822 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.752894 (* 0.3 = 0.225868 loss) | |
| I0330 07:13:11.441833 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404255 | |
| I0330 07:13:11.441846 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 07:13:11.441859 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.702128 | |
| I0330 07:13:11.441872 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.1382 (* 0.3 = 0.641461 loss) | |
| I0330 07:13:11.441887 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.624568 (* 0.3 = 0.18737 loss) | |
| I0330 07:13:11.441900 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.553191 | |
| I0330 07:13:11.441912 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0330 07:13:11.441925 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.765957 | |
| I0330 07:13:11.441939 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56797 (* 1 = 1.56797 loss) | |
| I0330 07:13:11.441952 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.439249 (* 1 = 0.439249 loss) | |
| I0330 07:13:11.441964 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:13:11.441977 10583 solver.cpp:245] Train net output #16: total_confidence = 0.209573 | |
| I0330 07:13:11.441989 10583 sgd_solver.cpp:106] Iteration 84500, lr = 0.01 | |
| I0330 07:15:20.749105 10583 solver.cpp:338] Iteration 85000, Testing net (#0) | |
| I0330 07:15:50.632815 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 07:15:50.632868 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 07:15:50.632884 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759047 | |
| I0330 07:15:50.632897 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 07:15:50.632913 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:15:50.632930 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:15:50.632942 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 07:15:50.632954 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759047 | |
| I0330 07:15:50.632966 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 07:15:50.632980 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:15:50.632994 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:15:50.633008 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 07:15:50.633019 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759047 | |
| I0330 07:15:50.633031 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 07:15:50.633045 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:15:50.633059 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:15:50.633072 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 07:15:50.633083 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 07:15:50.784394 10583 solver.cpp:229] Iteration 85000, loss = 3.51081 | |
| I0330 07:15:50.784484 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0330 07:15:50.784510 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 07:15:50.784523 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.444444 | |
| I0330 07:15:50.784538 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51825 (* 0.3 = 0.755476 loss) | |
| I0330 07:15:50.784554 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.714219 (* 0.3 = 0.214266 loss) | |
| I0330 07:15:50.784566 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.422222 | |
| I0330 07:15:50.784579 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 07:15:50.784590 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.577778 | |
| I0330 07:15:50.784603 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.01454 (* 0.3 = 0.604363 loss) | |
| I0330 07:15:50.784617 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.569029 (* 0.3 = 0.170709 loss) | |
| I0330 07:15:50.784629 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.733333 | |
| I0330 07:15:50.784646 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 07:15:50.784658 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0330 07:15:50.784672 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.978769 (* 1 = 0.978769 loss) | |
| I0330 07:15:50.784687 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.275453 (* 1 = 0.275453 loss) | |
| I0330 07:15:50.784698 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:15:50.784710 10583 solver.cpp:245] Train net output #16: total_confidence = 0.131448 | |
| I0330 07:15:50.784723 10583 sgd_solver.cpp:106] Iteration 85000, lr = 0.01 | |
| I0330 07:18:00.128751 10583 solver.cpp:229] Iteration 85500, loss = 3.5197 | |
| I0330 07:18:00.128888 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.317073 | |
| I0330 07:18:00.128911 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 07:18:00.128923 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.463415 | |
| I0330 07:18:00.128940 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.44467 (* 0.3 = 0.733401 loss) | |
| I0330 07:18:00.128955 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.670032 (* 0.3 = 0.20101 loss) | |
| I0330 07:18:00.128968 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.317073 | |
| I0330 07:18:00.128980 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 07:18:00.128993 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.560976 | |
| I0330 07:18:00.129006 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06242 (* 0.3 = 0.618726 loss) | |
| I0330 07:18:00.129020 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.588627 (* 0.3 = 0.176588 loss) | |
| I0330 07:18:00.129034 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.682927 | |
| I0330 07:18:00.129046 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 07:18:00.129058 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804878 | |
| I0330 07:18:00.129072 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14652 (* 1 = 1.14652 loss) | |
| I0330 07:18:00.129086 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.303661 (* 1 = 0.303661 loss) | |
| I0330 07:18:00.129099 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 07:18:00.129112 10583 solver.cpp:245] Train net output #16: total_confidence = 0.17964 | |
| I0330 07:18:00.129123 10583 sgd_solver.cpp:106] Iteration 85500, lr = 0.01 | |
| I0330 07:20:09.432499 10583 solver.cpp:229] Iteration 86000, loss = 3.521 | |
| I0330 07:20:09.432646 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.32 | |
| I0330 07:20:09.432667 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 07:20:09.432680 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0330 07:20:09.432698 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.16465 (* 0.3 = 0.649396 loss) | |
| I0330 07:20:09.432713 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.65144 (* 0.3 = 0.195432 loss) | |
| I0330 07:20:09.432731 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 07:20:09.432746 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 07:20:09.432760 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0330 07:20:09.432773 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.86196 (* 0.3 = 0.558589 loss) | |
| I0330 07:20:09.432788 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.555553 (* 0.3 = 0.166666 loss) | |
| I0330 07:20:09.432801 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.66 | |
| I0330 07:20:09.432812 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 07:20:09.432824 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 07:20:09.432839 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.19374 (* 1 = 1.19374 loss) | |
| I0330 07:20:09.432853 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.351865 (* 1 = 0.351865 loss) | |
| I0330 07:20:09.432865 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 07:20:09.432878 10583 solver.cpp:245] Train net output #16: total_confidence = 0.284222 | |
| I0330 07:20:09.432889 10583 sgd_solver.cpp:106] Iteration 86000, lr = 0.01 | |
| I0330 07:22:18.645917 10583 solver.cpp:229] Iteration 86500, loss = 3.49982 | |
| I0330 07:22:18.646044 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.434783 | |
| I0330 07:22:18.646064 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 07:22:18.646078 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.673913 | |
| I0330 07:22:18.646095 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.0105 (* 0.3 = 0.603151 loss) | |
| I0330 07:22:18.646109 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.596116 (* 0.3 = 0.178835 loss) | |
| I0330 07:22:18.646126 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.630435 | |
| I0330 07:22:18.646138 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 07:22:18.646152 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0330 07:22:18.646168 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67952 (* 0.3 = 0.503857 loss) | |
| I0330 07:22:18.646183 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.525599 (* 0.3 = 0.15768 loss) | |
| I0330 07:22:18.646195 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.717391 | |
| I0330 07:22:18.646215 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 07:22:18.646227 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0330 07:22:18.646242 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.954796 (* 1 = 0.954796 loss) | |
| I0330 07:22:18.646257 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.302268 (* 1 = 0.302268 loss) | |
| I0330 07:22:18.646270 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:22:18.646281 10583 solver.cpp:245] Train net output #16: total_confidence = 0.247568 | |
| I0330 07:22:18.646293 10583 sgd_solver.cpp:106] Iteration 86500, lr = 0.01 | |
| I0330 07:24:28.034477 10583 solver.cpp:229] Iteration 87000, loss = 3.41818 | |
| I0330 07:24:28.034682 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.264151 | |
| I0330 07:24:28.034703 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 07:24:28.034716 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.566038 | |
| I0330 07:24:28.034741 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39841 (* 0.3 = 0.719522 loss) | |
| I0330 07:24:28.034756 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.751571 (* 0.3 = 0.225471 loss) | |
| I0330 07:24:28.034770 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.320755 | |
| I0330 07:24:28.034781 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 07:24:28.034795 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.773585 | |
| I0330 07:24:28.034807 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.98893 (* 0.3 = 0.59668 loss) | |
| I0330 07:24:28.034828 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.625846 (* 0.3 = 0.187754 loss) | |
| I0330 07:24:28.034840 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.54717 | |
| I0330 07:24:28.034852 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 07:24:28.034864 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.867925 | |
| I0330 07:24:28.034888 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.35623 (* 1 = 1.35623 loss) | |
| I0330 07:24:28.034901 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.415717 (* 1 = 0.415717 loss) | |
| I0330 07:24:28.034914 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:24:28.034926 10583 solver.cpp:245] Train net output #16: total_confidence = 0.130705 | |
| I0330 07:24:28.034940 10583 sgd_solver.cpp:106] Iteration 87000, lr = 0.01 | |
| I0330 07:26:37.421497 10583 solver.cpp:229] Iteration 87500, loss = 3.42561 | |
| I0330 07:26:37.421613 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326087 | |
| I0330 07:26:37.421633 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 07:26:37.421646 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0330 07:26:37.421663 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07403 (* 0.3 = 0.622208 loss) | |
| I0330 07:26:37.421679 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.637309 (* 0.3 = 0.191193 loss) | |
| I0330 07:26:37.421691 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.478261 | |
| I0330 07:26:37.421703 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 07:26:37.421715 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.695652 | |
| I0330 07:26:37.421730 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.71857 (* 0.3 = 0.515572 loss) | |
| I0330 07:26:37.421744 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.526726 (* 0.3 = 0.158018 loss) | |
| I0330 07:26:37.421757 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.652174 | |
| I0330 07:26:37.421769 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 07:26:37.421782 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0330 07:26:37.421795 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.10899 (* 1 = 1.10899 loss) | |
| I0330 07:26:37.421809 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.377204 (* 1 = 0.377204 loss) | |
| I0330 07:26:37.421821 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:26:37.421833 10583 solver.cpp:245] Train net output #16: total_confidence = 0.226999 | |
| I0330 07:26:37.421845 10583 sgd_solver.cpp:106] Iteration 87500, lr = 0.01 | |
| I0330 07:28:46.886693 10583 solver.cpp:229] Iteration 88000, loss = 3.46411 | |
| I0330 07:28:46.886800 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 07:28:46.886828 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 07:28:46.886842 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.604167 | |
| I0330 07:28:46.886857 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.35028 (* 0.3 = 0.705083 loss) | |
| I0330 07:28:46.886873 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.735112 (* 0.3 = 0.220534 loss) | |
| I0330 07:28:46.886884 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.395833 | |
| I0330 07:28:46.886898 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 07:28:46.886912 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.645833 | |
| I0330 07:28:46.886935 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06813 (* 0.3 = 0.620439 loss) | |
| I0330 07:28:46.886951 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.633294 (* 0.3 = 0.189988 loss) | |
| I0330 07:28:46.886963 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0330 07:28:46.886988 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 07:28:46.887003 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 07:28:46.887017 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.30733 (* 1 = 1.30733 loss) | |
| I0330 07:28:46.887032 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.364749 (* 1 = 0.364749 loss) | |
| I0330 07:28:46.887044 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 07:28:46.887056 10583 solver.cpp:245] Train net output #16: total_confidence = 0.149162 | |
| I0330 07:28:46.887068 10583 sgd_solver.cpp:106] Iteration 88000, lr = 0.01 | |
| I0330 07:30:56.278849 10583 solver.cpp:229] Iteration 88500, loss = 3.46169 | |
| I0330 07:30:56.278997 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.342105 | |
| I0330 07:30:56.279019 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 07:30:56.279033 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.657895 | |
| I0330 07:30:56.279050 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15638 (* 0.3 = 0.646914 loss) | |
| I0330 07:30:56.279065 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.589137 (* 0.3 = 0.176741 loss) | |
| I0330 07:30:56.279078 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 07:30:56.279090 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 07:30:56.279103 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.736842 | |
| I0330 07:30:56.279117 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.70058 (* 0.3 = 0.510175 loss) | |
| I0330 07:30:56.279131 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.530671 (* 0.3 = 0.159201 loss) | |
| I0330 07:30:56.279145 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.763158 | |
| I0330 07:30:56.279156 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 07:30:56.279171 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.815789 | |
| I0330 07:30:56.279186 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.968537 (* 1 = 0.968537 loss) | |
| I0330 07:30:56.279201 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.266469 (* 1 = 0.266469 loss) | |
| I0330 07:30:56.279212 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 07:30:56.279224 10583 solver.cpp:245] Train net output #16: total_confidence = 0.308944 | |
| I0330 07:30:56.279237 10583 sgd_solver.cpp:106] Iteration 88500, lr = 0.01 | |
| I0330 07:33:05.723578 10583 solver.cpp:229] Iteration 89000, loss = 3.45407 | |
| I0330 07:33:05.723732 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.169811 | |
| I0330 07:33:05.723753 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0330 07:33:05.723767 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.339623 | |
| I0330 07:33:05.723783 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.7995 (* 0.3 = 0.83985 loss) | |
| I0330 07:33:05.723804 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.98834 (* 0.3 = 0.296502 loss) | |
| I0330 07:33:05.723816 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.207547 | |
| I0330 07:33:05.723829 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0330 07:33:05.723841 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.471698 | |
| I0330 07:33:05.723855 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.72113 (* 0.3 = 0.816338 loss) | |
| I0330 07:33:05.723870 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.977647 (* 0.3 = 0.293294 loss) | |
| I0330 07:33:05.723882 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.433962 | |
| I0330 07:33:05.723896 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0330 07:33:05.723907 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.716981 | |
| I0330 07:33:05.723922 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.83637 (* 1 = 1.83637 loss) | |
| I0330 07:33:05.723935 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.821956 (* 1 = 0.821956 loss) | |
| I0330 07:33:05.723948 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 07:33:05.723959 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0564708 | |
| I0330 07:33:05.723973 10583 sgd_solver.cpp:106] Iteration 89000, lr = 0.01 | |
| I0330 07:35:15.067811 10583 solver.cpp:229] Iteration 89500, loss = 3.45047 | |
| I0330 07:35:15.067929 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 07:35:15.067948 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 07:35:15.067961 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 07:35:15.067977 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34046 (* 0.3 = 0.702138 loss) | |
| I0330 07:35:15.067992 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.655646 (* 0.3 = 0.196694 loss) | |
| I0330 07:35:15.068004 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.466667 | |
| I0330 07:35:15.068017 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 07:35:15.068029 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 07:35:15.068043 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.78483 (* 0.3 = 0.535449 loss) | |
| I0330 07:35:15.068058 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.498977 (* 0.3 = 0.149693 loss) | |
| I0330 07:35:15.068069 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.555556 | |
| I0330 07:35:15.068081 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 07:35:15.068092 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0330 07:35:15.068106 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34175 (* 1 = 1.34175 loss) | |
| I0330 07:35:15.068122 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.354318 (* 1 = 0.354318 loss) | |
| I0330 07:35:15.068135 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:35:15.068146 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0922212 | |
| I0330 07:35:15.068161 10583 sgd_solver.cpp:106] Iteration 89500, lr = 0.01 | |
| I0330 07:37:24.286468 10583 solver.cpp:338] Iteration 90000, Testing net (#0) | |
| I0330 07:37:54.101903 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 07:37:54.101964 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 07:37:54.101980 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.75991 | |
| I0330 07:37:54.101994 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 07:37:54.102011 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:37:54.102027 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:37:54.102040 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 07:37:54.102052 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.75991 | |
| I0330 07:37:54.102064 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 07:37:54.102078 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:37:54.102093 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:37:54.102105 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 07:37:54.102118 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.75991 | |
| I0330 07:37:54.102129 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 07:37:54.102143 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:37:54.102157 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:37:54.102172 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 07:37:54.102185 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 07:37:54.253262 10583 solver.cpp:229] Iteration 90000, loss = 3.39431 | |
| I0330 07:37:54.253304 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.347826 | |
| I0330 07:37:54.253320 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 07:37:54.253334 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.608696 | |
| I0330 07:37:54.253348 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.05895 (* 0.3 = 0.617684 loss) | |
| I0330 07:37:54.253363 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.669899 (* 0.3 = 0.20097 loss) | |
| I0330 07:37:54.253376 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.413043 | |
| I0330 07:37:54.253388 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 07:37:54.253399 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.804348 | |
| I0330 07:37:54.253413 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.85261 (* 0.3 = 0.555783 loss) | |
| I0330 07:37:54.253428 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.618261 (* 0.3 = 0.185478 loss) | |
| I0330 07:37:54.253440 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.782609 | |
| I0330 07:37:54.253453 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 07:37:54.253464 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.913043 | |
| I0330 07:37:54.253479 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.895221 (* 1 = 0.895221 loss) | |
| I0330 07:37:54.253492 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.290335 (* 1 = 0.290335 loss) | |
| I0330 07:37:54.253504 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:37:54.253517 10583 solver.cpp:245] Train net output #16: total_confidence = 0.231983 | |
| I0330 07:37:54.253530 10583 sgd_solver.cpp:106] Iteration 90000, lr = 0.01 | |
| I0330 07:40:03.451303 10583 solver.cpp:229] Iteration 90500, loss = 3.40874 | |
| I0330 07:40:03.451455 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0330 07:40:03.451475 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 07:40:03.451488 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.413043 | |
| I0330 07:40:03.451505 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4946 (* 0.3 = 0.748379 loss) | |
| I0330 07:40:03.451520 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.689208 (* 0.3 = 0.206762 loss) | |
| I0330 07:40:03.451534 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.326087 | |
| I0330 07:40:03.451553 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 07:40:03.451565 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.586957 | |
| I0330 07:40:03.451580 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04197 (* 0.3 = 0.612592 loss) | |
| I0330 07:40:03.451593 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.593687 (* 0.3 = 0.178106 loss) | |
| I0330 07:40:03.451606 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.608696 | |
| I0330 07:40:03.451619 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0330 07:40:03.451632 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.73913 | |
| I0330 07:40:03.451645 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3342 (* 1 = 1.3342 loss) | |
| I0330 07:40:03.451659 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.4116 (* 1 = 0.4116 loss) | |
| I0330 07:40:03.451671 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 07:40:03.451683 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0510277 | |
| I0330 07:40:03.451696 10583 sgd_solver.cpp:106] Iteration 90500, lr = 0.01 | |
| I0330 07:42:12.746366 10583 solver.cpp:229] Iteration 91000, loss = 3.42258 | |
| I0330 07:42:12.746490 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.234043 | |
| I0330 07:42:12.746510 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 07:42:12.746523 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.531915 | |
| I0330 07:42:12.746541 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.24103 (* 0.3 = 0.672308 loss) | |
| I0330 07:42:12.746556 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.661905 (* 0.3 = 0.198571 loss) | |
| I0330 07:42:12.746567 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.425532 | |
| I0330 07:42:12.746580 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 07:42:12.746592 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.617021 | |
| I0330 07:42:12.746605 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83448 (* 0.3 = 0.550344 loss) | |
| I0330 07:42:12.746619 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.536904 (* 0.3 = 0.161071 loss) | |
| I0330 07:42:12.746633 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.659574 | |
| I0330 07:42:12.746644 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 07:42:12.746655 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.87234 | |
| I0330 07:42:12.746670 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.06909 (* 1 = 1.06909 loss) | |
| I0330 07:42:12.746683 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.29875 (* 1 = 0.29875 loss) | |
| I0330 07:42:12.746695 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:42:12.746707 10583 solver.cpp:245] Train net output #16: total_confidence = 0.158953 | |
| I0330 07:42:12.746719 10583 sgd_solver.cpp:106] Iteration 91000, lr = 0.01 | |
| I0330 07:44:22.215322 10583 solver.cpp:229] Iteration 91500, loss = 3.34069 | |
| I0330 07:44:22.215461 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.377778 | |
| I0330 07:44:22.215481 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 07:44:22.215493 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.711111 | |
| I0330 07:44:22.215510 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.00763 (* 0.3 = 0.602289 loss) | |
| I0330 07:44:22.215533 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.600221 (* 0.3 = 0.180066 loss) | |
| I0330 07:44:22.215545 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.444444 | |
| I0330 07:44:22.215558 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 07:44:22.215569 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 07:44:22.215584 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.55929 (* 0.3 = 0.467786 loss) | |
| I0330 07:44:22.215598 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.478061 (* 0.3 = 0.143418 loss) | |
| I0330 07:44:22.215610 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0330 07:44:22.215622 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 07:44:22.215634 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977778 | |
| I0330 07:44:22.215648 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.668211 (* 1 = 0.668211 loss) | |
| I0330 07:44:22.215662 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.221059 (* 1 = 0.221059 loss) | |
| I0330 07:44:22.215677 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:44:22.215690 10583 solver.cpp:245] Train net output #16: total_confidence = 0.201269 | |
| I0330 07:44:22.215703 10583 sgd_solver.cpp:106] Iteration 91500, lr = 0.01 | |
| I0330 07:46:31.628916 10583 solver.cpp:229] Iteration 92000, loss = 3.37729 | |
| I0330 07:46:31.629022 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.37037 | |
| I0330 07:46:31.629042 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 07:46:31.629055 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.611111 | |
| I0330 07:46:31.629072 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.1178 (* 0.3 = 0.635339 loss) | |
| I0330 07:46:31.629087 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.674721 (* 0.3 = 0.202416 loss) | |
| I0330 07:46:31.629101 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 07:46:31.629112 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 07:46:31.629124 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.722222 | |
| I0330 07:46:31.629138 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9473 (* 0.3 = 0.584191 loss) | |
| I0330 07:46:31.629153 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.61028 (* 0.3 = 0.183084 loss) | |
| I0330 07:46:31.629164 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 07:46:31.629176 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 07:46:31.629189 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 07:46:31.629202 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.0631 (* 1 = 1.0631 loss) | |
| I0330 07:46:31.629217 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.336378 (* 1 = 0.336378 loss) | |
| I0330 07:46:31.629230 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 07:46:31.629241 10583 solver.cpp:245] Train net output #16: total_confidence = 0.150531 | |
| I0330 07:46:31.629253 10583 sgd_solver.cpp:106] Iteration 92000, lr = 0.01 | |
| I0330 07:48:41.772665 10583 solver.cpp:229] Iteration 92500, loss = 3.3735 | |
| I0330 07:48:41.772781 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.294118 | |
| I0330 07:48:41.772810 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 07:48:41.772824 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.529412 | |
| I0330 07:48:41.772840 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34821 (* 0.3 = 0.704464 loss) | |
| I0330 07:48:41.772855 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.711219 (* 0.3 = 0.213366 loss) | |
| I0330 07:48:41.772867 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.470588 | |
| I0330 07:48:41.772881 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 07:48:41.772893 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.764706 | |
| I0330 07:48:41.772907 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68703 (* 0.3 = 0.506108 loss) | |
| I0330 07:48:41.772922 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.549262 (* 0.3 = 0.164779 loss) | |
| I0330 07:48:41.772934 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 07:48:41.772946 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 07:48:41.772958 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.823529 | |
| I0330 07:48:41.772972 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.0076 (* 1 = 1.0076 loss) | |
| I0330 07:48:41.772986 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.344242 (* 1 = 0.344242 loss) | |
| I0330 07:48:41.772999 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 07:48:41.773010 10583 solver.cpp:245] Train net output #16: total_confidence = 0.20193 | |
| I0330 07:48:41.773023 10583 sgd_solver.cpp:106] Iteration 92500, lr = 0.01 | |
| I0330 07:50:51.282467 10583 solver.cpp:229] Iteration 93000, loss = 3.31255 | |
| I0330 07:50:51.282582 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372549 | |
| I0330 07:50:51.282603 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 07:50:51.282615 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.588235 | |
| I0330 07:50:51.282631 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23502 (* 0.3 = 0.670507 loss) | |
| I0330 07:50:51.282655 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.663404 (* 0.3 = 0.199021 loss) | |
| I0330 07:50:51.282667 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431373 | |
| I0330 07:50:51.282680 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 07:50:51.282691 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.705882 | |
| I0330 07:50:51.282712 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.89317 (* 0.3 = 0.56795 loss) | |
| I0330 07:50:51.282727 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.562057 (* 0.3 = 0.168617 loss) | |
| I0330 07:50:51.282738 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 07:50:51.282750 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 07:50:51.282763 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.843137 | |
| I0330 07:50:51.282776 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.956481 (* 1 = 0.956481 loss) | |
| I0330 07:50:51.282790 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.306403 (* 1 = 0.306403 loss) | |
| I0330 07:50:51.282802 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:50:51.282814 10583 solver.cpp:245] Train net output #16: total_confidence = 0.199814 | |
| I0330 07:50:51.282835 10583 sgd_solver.cpp:106] Iteration 93000, lr = 0.01 | |
| I0330 07:53:00.804044 10583 solver.cpp:229] Iteration 93500, loss = 3.37961 | |
| I0330 07:53:00.804225 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.304348 | |
| I0330 07:53:00.804246 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 07:53:00.804260 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.521739 | |
| I0330 07:53:00.804275 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.21745 (* 0.3 = 0.665235 loss) | |
| I0330 07:53:00.804291 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.611624 (* 0.3 = 0.183487 loss) | |
| I0330 07:53:00.804312 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.456522 | |
| I0330 07:53:00.804325 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 07:53:00.804337 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.608696 | |
| I0330 07:53:00.804350 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.90541 (* 0.3 = 0.571622 loss) | |
| I0330 07:53:00.804365 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524368 (* 0.3 = 0.15731 loss) | |
| I0330 07:53:00.804378 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.608696 | |
| I0330 07:53:00.804390 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 07:53:00.804401 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0330 07:53:00.804416 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29627 (* 1 = 1.29627 loss) | |
| I0330 07:53:00.804430 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.3723 (* 1 = 0.3723 loss) | |
| I0330 07:53:00.804443 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:53:00.804455 10583 solver.cpp:245] Train net output #16: total_confidence = 0.289157 | |
| I0330 07:53:00.804469 10583 sgd_solver.cpp:106] Iteration 93500, lr = 0.01 | |
| I0330 07:55:09.881780 10583 solver.cpp:229] Iteration 94000, loss = 3.33499 | |
| I0330 07:55:09.881909 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0330 07:55:09.881929 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 07:55:09.881942 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 07:55:09.881958 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64847 (* 0.3 = 0.794542 loss) | |
| I0330 07:55:09.881973 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.79185 (* 0.3 = 0.237555 loss) | |
| I0330 07:55:09.881986 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.26 | |
| I0330 07:55:09.881997 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 07:55:09.882009 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.54 | |
| I0330 07:55:09.882024 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.55285 (* 0.3 = 0.765855 loss) | |
| I0330 07:55:09.882037 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.746753 (* 0.3 = 0.224026 loss) | |
| I0330 07:55:09.882050 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.44 | |
| I0330 07:55:09.882062 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 07:55:09.882073 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.64 | |
| I0330 07:55:09.882087 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.12351 (* 1 = 2.12351 loss) | |
| I0330 07:55:09.882102 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.619136 (* 1 = 0.619136 loss) | |
| I0330 07:55:09.882114 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:55:09.882127 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0522579 | |
| I0330 07:55:09.882138 10583 sgd_solver.cpp:106] Iteration 94000, lr = 0.01 | |
| I0330 07:57:19.391522 10583 solver.cpp:229] Iteration 94500, loss = 3.37973 | |
| I0330 07:57:19.391666 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.325 | |
| I0330 07:57:19.391686 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 07:57:19.391700 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.675 | |
| I0330 07:57:19.391716 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80714 (* 0.3 = 0.542143 loss) | |
| I0330 07:57:19.391737 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.490697 (* 0.3 = 0.147209 loss) | |
| I0330 07:57:19.391751 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 07:57:19.391762 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 07:57:19.391774 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.725 | |
| I0330 07:57:19.391788 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.58753 (* 0.3 = 0.476259 loss) | |
| I0330 07:57:19.391803 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.508837 (* 0.3 = 0.152651 loss) | |
| I0330 07:57:19.391816 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0330 07:57:19.391829 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 07:57:19.391840 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0330 07:57:19.391855 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.993302 (* 1 = 0.993302 loss) | |
| I0330 07:57:19.391868 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.271794 (* 1 = 0.271794 loss) | |
| I0330 07:57:19.391881 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:57:19.391893 10583 solver.cpp:245] Train net output #16: total_confidence = 0.251485 | |
| I0330 07:57:19.391906 10583 sgd_solver.cpp:106] Iteration 94500, lr = 0.01 | |
| I0330 07:59:28.547971 10583 solver.cpp:338] Iteration 95000, Testing net (#0) | |
| I0330 07:59:58.391762 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 07:59:58.391810 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 07:59:58.391826 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759637 | |
| I0330 07:59:58.391839 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 07:59:58.391856 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:59:58.391871 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:59:58.391885 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 07:59:58.391896 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759637 | |
| I0330 07:59:58.391907 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 07:59:58.391921 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:59:58.391937 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 07:59:58.391948 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 07:59:58.391960 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759637 | |
| I0330 07:59:58.391973 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 07:59:58.391986 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:59:58.392001 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 07:59:58.392014 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 07:59:58.392025 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 07:59:58.543524 10583 solver.cpp:229] Iteration 95000, loss = 3.28603 | |
| I0330 07:59:58.543565 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.394737 | |
| I0330 07:59:58.543582 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 07:59:58.543594 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.578947 | |
| I0330 07:59:58.543609 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14892 (* 0.3 = 0.644675 loss) | |
| I0330 07:59:58.543624 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.578572 (* 0.3 = 0.173572 loss) | |
| I0330 07:59:58.543637 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.605263 | |
| I0330 07:59:58.543648 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 07:59:58.543660 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.868421 | |
| I0330 07:59:58.543674 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39069 (* 0.3 = 0.417207 loss) | |
| I0330 07:59:58.543689 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.385657 (* 0.3 = 0.115697 loss) | |
| I0330 07:59:58.543701 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.657895 | |
| I0330 07:59:58.543714 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 07:59:58.543727 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.921053 | |
| I0330 07:59:58.543742 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.888495 (* 1 = 0.888495 loss) | |
| I0330 07:59:58.543756 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.228733 (* 1 = 0.228733 loss) | |
| I0330 07:59:58.543768 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 07:59:58.543781 10583 solver.cpp:245] Train net output #16: total_confidence = 0.26384 | |
| I0330 07:59:58.543794 10583 sgd_solver.cpp:106] Iteration 95000, lr = 0.01 | |
| I0330 08:02:07.850656 10583 solver.cpp:229] Iteration 95500, loss = 3.30961 | |
| I0330 08:02:07.850805 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.204545 | |
| I0330 08:02:07.850826 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 08:02:07.850839 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.545455 | |
| I0330 08:02:07.850855 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64816 (* 0.3 = 0.794447 loss) | |
| I0330 08:02:07.850870 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.729078 (* 0.3 = 0.218723 loss) | |
| I0330 08:02:07.850890 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.295455 | |
| I0330 08:02:07.850903 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 08:02:07.850915 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.545455 | |
| I0330 08:02:07.850929 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.4324 (* 0.3 = 0.729721 loss) | |
| I0330 08:02:07.850944 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.687861 (* 0.3 = 0.206358 loss) | |
| I0330 08:02:07.850956 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.454545 | |
| I0330 08:02:07.850968 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 08:02:07.850993 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.704545 | |
| I0330 08:02:07.851011 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.90988 (* 1 = 1.90988 loss) | |
| I0330 08:02:07.851024 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.549045 (* 1 = 0.549045 loss) | |
| I0330 08:02:07.851037 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 08:02:07.851049 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0890806 | |
| I0330 08:02:07.851063 10583 sgd_solver.cpp:106] Iteration 95500, lr = 0.01 | |
| I0330 08:04:17.098629 10583 solver.cpp:229] Iteration 96000, loss = 3.34876 | |
| I0330 08:04:17.098773 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.234043 | |
| I0330 08:04:17.098794 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 08:04:17.098808 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510638 | |
| I0330 08:04:17.098832 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46308 (* 0.3 = 0.738925 loss) | |
| I0330 08:04:17.098847 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.7079 (* 0.3 = 0.21237 loss) | |
| I0330 08:04:17.098860 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.446809 | |
| I0330 08:04:17.098872 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 08:04:17.098884 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.595745 | |
| I0330 08:04:17.098898 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12309 (* 0.3 = 0.636927 loss) | |
| I0330 08:04:17.098912 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.630743 (* 0.3 = 0.189223 loss) | |
| I0330 08:04:17.098924 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.574468 | |
| I0330 08:04:17.098937 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 08:04:17.098948 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.744681 | |
| I0330 08:04:17.098963 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.66475 (* 1 = 1.66475 loss) | |
| I0330 08:04:17.098989 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.463752 (* 1 = 0.463752 loss) | |
| I0330 08:04:17.099004 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:04:17.099016 10583 solver.cpp:245] Train net output #16: total_confidence = 0.138458 | |
| I0330 08:04:17.099028 10583 sgd_solver.cpp:106] Iteration 96000, lr = 0.01 | |
| I0330 08:06:26.414026 10583 solver.cpp:229] Iteration 96500, loss = 3.30012 | |
| I0330 08:06:26.414135 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0330 08:06:26.414155 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 08:06:26.414170 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0330 08:06:26.414188 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36866 (* 0.3 = 0.710597 loss) | |
| I0330 08:06:26.414203 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.698279 (* 0.3 = 0.209484 loss) | |
| I0330 08:06:26.414216 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.456522 | |
| I0330 08:06:26.414227 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 08:06:26.414239 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.695652 | |
| I0330 08:06:26.414253 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.92024 (* 0.3 = 0.576073 loss) | |
| I0330 08:06:26.414268 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.563901 (* 0.3 = 0.16917 loss) | |
| I0330 08:06:26.414279 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.695652 | |
| I0330 08:06:26.414291 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 08:06:26.414304 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826087 | |
| I0330 08:06:26.414317 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.08118 (* 1 = 1.08118 loss) | |
| I0330 08:06:26.414331 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.329415 (* 1 = 0.329415 loss) | |
| I0330 08:06:26.414345 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:06:26.414356 10583 solver.cpp:245] Train net output #16: total_confidence = 0.149983 | |
| I0330 08:06:26.414368 10583 sgd_solver.cpp:106] Iteration 96500, lr = 0.01 | |
| I0330 08:08:35.902529 10583 solver.cpp:229] Iteration 97000, loss = 3.22115 | |
| I0330 08:08:35.902676 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.325581 | |
| I0330 08:08:35.902698 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 08:08:35.902710 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.581395 | |
| I0330 08:08:35.902726 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.125 (* 0.3 = 0.6375 loss) | |
| I0330 08:08:35.902750 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.570022 (* 0.3 = 0.171007 loss) | |
| I0330 08:08:35.902761 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.465116 | |
| I0330 08:08:35.902775 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 08:08:35.902786 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744186 | |
| I0330 08:08:35.902801 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.57652 (* 0.3 = 0.472956 loss) | |
| I0330 08:08:35.902815 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.441261 (* 0.3 = 0.132378 loss) | |
| I0330 08:08:35.902832 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.651163 | |
| I0330 08:08:35.902844 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 08:08:35.902855 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0330 08:08:35.902870 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.972612 (* 1 = 0.972612 loss) | |
| I0330 08:08:35.902891 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.274216 (* 1 = 0.274216 loss) | |
| I0330 08:08:35.902904 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:08:35.902916 10583 solver.cpp:245] Train net output #16: total_confidence = 0.27439 | |
| I0330 08:08:35.902928 10583 sgd_solver.cpp:106] Iteration 97000, lr = 0.01 | |
| I0330 08:10:45.417356 10583 solver.cpp:229] Iteration 97500, loss = 3.28615 | |
| I0330 08:10:45.417479 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0330 08:10:45.417500 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 08:10:45.417512 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.541667 | |
| I0330 08:10:45.417528 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41859 (* 0.3 = 0.725578 loss) | |
| I0330 08:10:45.417543 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.69574 (* 0.3 = 0.208722 loss) | |
| I0330 08:10:45.417557 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.354167 | |
| I0330 08:10:45.417568 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 08:10:45.417582 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.708333 | |
| I0330 08:10:45.417595 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.87542 (* 0.3 = 0.562627 loss) | |
| I0330 08:10:45.417609 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.558433 (* 0.3 = 0.16753 loss) | |
| I0330 08:10:45.417621 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0330 08:10:45.417634 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 08:10:45.417645 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8125 | |
| I0330 08:10:45.417659 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24166 (* 1 = 1.24166 loss) | |
| I0330 08:10:45.417675 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.367493 (* 1 = 0.367493 loss) | |
| I0330 08:10:45.417686 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:10:45.417698 10583 solver.cpp:245] Train net output #16: total_confidence = 0.152851 | |
| I0330 08:10:45.417711 10583 sgd_solver.cpp:106] Iteration 97500, lr = 0.01 | |
| I0330 08:12:54.900352 10583 solver.cpp:229] Iteration 98000, loss = 3.29196 | |
| I0330 08:12:54.900501 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.276596 | |
| I0330 08:12:54.900521 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 08:12:54.900534 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.446809 | |
| I0330 08:12:54.900552 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4669 (* 0.3 = 0.740071 loss) | |
| I0330 08:12:54.900574 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.751005 (* 0.3 = 0.225302 loss) | |
| I0330 08:12:54.900586 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.319149 | |
| I0330 08:12:54.900599 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 08:12:54.900610 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.595745 | |
| I0330 08:12:54.900624 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.00895 (* 0.3 = 0.602686 loss) | |
| I0330 08:12:54.900640 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.62716 (* 0.3 = 0.188148 loss) | |
| I0330 08:12:54.900651 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.574468 | |
| I0330 08:12:54.900663 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 08:12:54.900676 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.702128 | |
| I0330 08:12:54.900689 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.48992 (* 1 = 1.48992 loss) | |
| I0330 08:12:54.900703 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.468394 (* 1 = 0.468394 loss) | |
| I0330 08:12:54.900717 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 08:12:54.900728 10583 solver.cpp:245] Train net output #16: total_confidence = 0.16402 | |
| I0330 08:12:54.900740 10583 sgd_solver.cpp:106] Iteration 98000, lr = 0.01 | |
| I0330 08:15:04.267506 10583 solver.cpp:229] Iteration 98500, loss = 3.25098 | |
| I0330 08:15:04.267628 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.377778 | |
| I0330 08:15:04.267647 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 08:15:04.267662 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.622222 | |
| I0330 08:15:04.267680 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32372 (* 0.3 = 0.697117 loss) | |
| I0330 08:15:04.267696 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.641078 (* 0.3 = 0.192323 loss) | |
| I0330 08:15:04.267709 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 08:15:04.267721 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 08:15:04.267734 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0330 08:15:04.267747 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.17426 (* 0.3 = 0.652278 loss) | |
| I0330 08:15:04.267761 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.613009 (* 0.3 = 0.183903 loss) | |
| I0330 08:15:04.267774 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.511111 | |
| I0330 08:15:04.267786 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 08:15:04.267798 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.688889 | |
| I0330 08:15:04.267812 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82693 (* 1 = 1.82693 loss) | |
| I0330 08:15:04.267827 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.516586 (* 1 = 0.516586 loss) | |
| I0330 08:15:04.267838 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:15:04.267850 10583 solver.cpp:245] Train net output #16: total_confidence = 0.206596 | |
| I0330 08:15:04.267863 10583 sgd_solver.cpp:106] Iteration 98500, lr = 0.01 | |
| I0330 08:17:13.586498 10583 solver.cpp:229] Iteration 99000, loss = 3.30609 | |
| I0330 08:17:13.586663 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.14 | |
| I0330 08:17:13.586683 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 08:17:13.586695 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.42 | |
| I0330 08:17:13.586719 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.78678 (* 0.3 = 0.836035 loss) | |
| I0330 08:17:13.586733 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.831325 (* 0.3 = 0.249397 loss) | |
| I0330 08:17:13.586746 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.24 | |
| I0330 08:17:13.586758 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 08:17:13.586771 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.48 | |
| I0330 08:17:13.586786 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.47123 (* 0.3 = 0.741368 loss) | |
| I0330 08:17:13.586799 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.762144 (* 0.3 = 0.228643 loss) | |
| I0330 08:17:13.586812 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.44 | |
| I0330 08:17:13.586829 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 08:17:13.586841 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.68 | |
| I0330 08:17:13.586855 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.81771 (* 1 = 1.81771 loss) | |
| I0330 08:17:13.586869 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.548229 (* 1 = 0.548229 loss) | |
| I0330 08:17:13.586887 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 08:17:13.586899 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0708238 | |
| I0330 08:17:13.586911 10583 sgd_solver.cpp:106] Iteration 99000, lr = 0.01 | |
| I0330 08:19:22.889313 10583 solver.cpp:229] Iteration 99500, loss = 3.23409 | |
| I0330 08:19:22.889418 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26 | |
| I0330 08:19:22.889438 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 08:19:22.889451 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 08:19:22.889468 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.75857 (* 0.3 = 0.827571 loss) | |
| I0330 08:19:22.889483 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.856388 (* 0.3 = 0.256916 loss) | |
| I0330 08:19:22.889495 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22 | |
| I0330 08:19:22.889508 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0330 08:19:22.889520 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0330 08:19:22.889534 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.26157 (* 0.3 = 0.678472 loss) | |
| I0330 08:19:22.889549 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.697186 (* 0.3 = 0.209156 loss) | |
| I0330 08:19:22.889560 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.54 | |
| I0330 08:19:22.889572 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0330 08:19:22.889585 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 08:19:22.889598 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.48823 (* 1 = 1.48823 loss) | |
| I0330 08:19:22.889612 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.493747 (* 1 = 0.493747 loss) | |
| I0330 08:19:22.889626 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 08:19:22.889636 10583 solver.cpp:245] Train net output #16: total_confidence = 0.131888 | |
| I0330 08:19:22.889649 10583 sgd_solver.cpp:106] Iteration 99500, lr = 0.01 | |
| I0330 08:21:31.953129 10583 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_100000.caffemodel | |
| I0330 08:21:32.266530 10583 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_100000.solverstate | |
| I0330 08:21:32.434447 10583 solver.cpp:338] Iteration 100000, Testing net (#0) | |
| I0330 08:22:02.217816 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 08:22:02.217919 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 08:22:02.217939 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759318 | |
| I0330 08:22:02.217952 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 08:22:02.217969 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:22:02.217985 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:22:02.217998 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 08:22:02.218009 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759318 | |
| I0330 08:22:02.218022 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 08:22:02.218036 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:22:02.218050 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:22:02.218063 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 08:22:02.218075 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759318 | |
| I0330 08:22:02.218086 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 08:22:02.218101 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 08:22:02.218116 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 08:22:02.218127 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 08:22:02.218138 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 08:22:02.370081 10583 solver.cpp:229] Iteration 100000, loss = 3.23785 | |
| I0330 08:22:02.370141 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.265306 | |
| I0330 08:22:02.370158 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 08:22:02.370172 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0330 08:22:02.370189 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51503 (* 0.3 = 0.754508 loss) | |
| I0330 08:22:02.370203 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.733326 (* 0.3 = 0.219998 loss) | |
| I0330 08:22:02.370216 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.387755 | |
| I0330 08:22:02.370229 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 08:22:02.370241 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.632653 | |
| I0330 08:22:02.370255 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.15932 (* 0.3 = 0.647796 loss) | |
| I0330 08:22:02.370270 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.654118 (* 0.3 = 0.196235 loss) | |
| I0330 08:22:02.370282 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673469 | |
| I0330 08:22:02.370295 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 08:22:02.370306 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.836735 | |
| I0330 08:22:02.370321 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.20421 (* 1 = 1.20421 loss) | |
| I0330 08:22:02.370334 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.350995 (* 1 = 0.350995 loss) | |
| I0330 08:22:02.370347 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:22:02.370359 10583 solver.cpp:245] Train net output #16: total_confidence = 0.19796 | |
| I0330 08:22:02.370371 10583 sgd_solver.cpp:106] Iteration 100000, lr = 0.01 | |
| I0330 08:24:11.814810 10583 solver.cpp:229] Iteration 100500, loss = 3.26873 | |
| I0330 08:24:11.814965 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.313726 | |
| I0330 08:24:11.814985 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 08:24:11.814997 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.568627 | |
| I0330 08:24:11.815014 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.484 (* 0.3 = 0.745201 loss) | |
| I0330 08:24:11.815029 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.753988 (* 0.3 = 0.226196 loss) | |
| I0330 08:24:11.815050 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 08:24:11.815062 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 08:24:11.815074 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.627451 | |
| I0330 08:24:11.815088 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.08869 (* 0.3 = 0.626607 loss) | |
| I0330 08:24:11.815102 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.652284 (* 0.3 = 0.195685 loss) | |
| I0330 08:24:11.815115 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 08:24:11.815127 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 08:24:11.815140 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.823529 | |
| I0330 08:24:11.815153 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.32852 (* 1 = 1.32852 loss) | |
| I0330 08:24:11.815171 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.425639 (* 1 = 0.425639 loss) | |
| I0330 08:24:11.815182 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:24:11.815196 10583 solver.cpp:245] Train net output #16: total_confidence = 0.144595 | |
| I0330 08:24:11.815207 10583 sgd_solver.cpp:106] Iteration 100500, lr = 0.01 | |
| I0330 08:26:21.065829 10583 solver.cpp:229] Iteration 101000, loss = 3.2717 | |
| I0330 08:26:21.065939 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.422222 | |
| I0330 08:26:21.065958 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 08:26:21.065971 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.733333 | |
| I0330 08:26:21.065989 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73834 (* 0.3 = 0.521503 loss) | |
| I0330 08:26:21.066004 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.471432 (* 0.3 = 0.14143 loss) | |
| I0330 08:26:21.066015 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0330 08:26:21.066028 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 08:26:21.066040 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0330 08:26:21.066054 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.37981 (* 0.3 = 0.413943 loss) | |
| I0330 08:26:21.066068 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.402901 (* 0.3 = 0.12087 loss) | |
| I0330 08:26:21.066082 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0330 08:26:21.066093 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 08:26:21.066105 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 08:26:21.066119 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.401908 (* 1 = 0.401908 loss) | |
| I0330 08:26:21.066133 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.128453 (* 1 = 0.128453 loss) | |
| I0330 08:26:21.066145 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:26:21.066160 10583 solver.cpp:245] Train net output #16: total_confidence = 0.296583 | |
| I0330 08:26:21.066174 10583 sgd_solver.cpp:106] Iteration 101000, lr = 0.01 | |
| I0330 08:28:30.439609 10583 solver.cpp:229] Iteration 101500, loss = 3.18562 | |
| I0330 08:28:30.439826 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.295455 | |
| I0330 08:28:30.439859 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 08:28:30.439893 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.613636 | |
| I0330 08:28:30.439926 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.27401 (* 0.3 = 0.682204 loss) | |
| I0330 08:28:30.439956 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.615827 (* 0.3 = 0.184748 loss) | |
| I0330 08:28:30.439985 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 08:28:30.440009 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 08:28:30.440037 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0330 08:28:30.440067 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.80039 (* 0.3 = 0.540116 loss) | |
| I0330 08:28:30.440098 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.487053 (* 0.3 = 0.146116 loss) | |
| I0330 08:28:30.440124 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.795455 | |
| I0330 08:28:30.440150 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 08:28:30.440174 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0330 08:28:30.440192 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.704621 (* 1 = 0.704621 loss) | |
| I0330 08:28:30.440207 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.193605 (* 1 = 0.193605 loss) | |
| I0330 08:28:30.440220 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 08:28:30.440232 10583 solver.cpp:245] Train net output #16: total_confidence = 0.205813 | |
| I0330 08:28:30.440245 10583 sgd_solver.cpp:106] Iteration 101500, lr = 0.01 | |
| I0330 08:30:39.596753 10583 solver.cpp:229] Iteration 102000, loss = 3.20342 | |
| I0330 08:30:39.596904 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0330 08:30:39.596925 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 08:30:39.596946 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595745 | |
| I0330 08:30:39.596962 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.28557 (* 0.3 = 0.685672 loss) | |
| I0330 08:30:39.596976 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.677016 (* 0.3 = 0.203105 loss) | |
| I0330 08:30:39.596989 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.446809 | |
| I0330 08:30:39.597002 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 08:30:39.597014 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744681 | |
| I0330 08:30:39.597028 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.91115 (* 0.3 = 0.573345 loss) | |
| I0330 08:30:39.597043 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.563755 (* 0.3 = 0.169126 loss) | |
| I0330 08:30:39.597055 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.787234 | |
| I0330 08:30:39.597067 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 08:30:39.597079 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.93617 | |
| I0330 08:30:39.597093 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.727033 (* 1 = 0.727033 loss) | |
| I0330 08:30:39.597108 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.209632 (* 1 = 0.209632 loss) | |
| I0330 08:30:39.597121 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:30:39.597132 10583 solver.cpp:245] Train net output #16: total_confidence = 0.23986 | |
| I0330 08:30:39.597146 10583 sgd_solver.cpp:106] Iteration 102000, lr = 0.01 | |
| I0330 08:32:49.124199 10583 solver.cpp:229] Iteration 102500, loss = 3.26005 | |
| I0330 08:32:49.124413 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 08:32:49.124433 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 08:32:49.124455 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.571429 | |
| I0330 08:32:49.124472 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.18399 (* 0.3 = 0.655196 loss) | |
| I0330 08:32:49.124487 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.647156 (* 0.3 = 0.194147 loss) | |
| I0330 08:32:49.124500 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.428571 | |
| I0330 08:32:49.124513 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 08:32:49.124526 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.642857 | |
| I0330 08:32:49.124541 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83142 (* 0.3 = 0.549425 loss) | |
| I0330 08:32:49.124554 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.560845 (* 0.3 = 0.168254 loss) | |
| I0330 08:32:49.124567 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0330 08:32:49.124578 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 08:32:49.124590 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.761905 | |
| I0330 08:32:49.124605 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26391 (* 1 = 1.26391 loss) | |
| I0330 08:32:49.124619 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.341114 (* 1 = 0.341114 loss) | |
| I0330 08:32:49.124631 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:32:49.124644 10583 solver.cpp:245] Train net output #16: total_confidence = 0.24367 | |
| I0330 08:32:49.124656 10583 sgd_solver.cpp:106] Iteration 102500, lr = 0.01 | |
| I0330 08:34:58.938305 10583 solver.cpp:229] Iteration 103000, loss = 3.14985 | |
| I0330 08:34:58.938474 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 08:34:58.938496 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 08:34:58.938509 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 08:34:58.938526 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11251 (* 0.3 = 0.633753 loss) | |
| I0330 08:34:58.938541 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.590782 (* 0.3 = 0.177235 loss) | |
| I0330 08:34:58.938555 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55 | |
| I0330 08:34:58.938567 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 08:34:58.938580 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.775 | |
| I0330 08:34:58.938594 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61861 (* 0.3 = 0.485584 loss) | |
| I0330 08:34:58.938608 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.497487 (* 0.3 = 0.149246 loss) | |
| I0330 08:34:58.938621 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.775 | |
| I0330 08:34:58.938634 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 08:34:58.938647 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.925 | |
| I0330 08:34:58.938660 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.677853 (* 1 = 0.677853 loss) | |
| I0330 08:34:58.938674 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.175126 (* 1 = 0.175126 loss) | |
| I0330 08:34:58.938688 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 08:34:58.938699 10583 solver.cpp:245] Train net output #16: total_confidence = 0.273623 | |
| I0330 08:34:58.938712 10583 sgd_solver.cpp:106] Iteration 103000, lr = 0.01 | |
| I0330 08:37:08.723268 10583 solver.cpp:229] Iteration 103500, loss = 3.20998 | |
| I0330 08:37:08.723438 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0330 08:37:08.723459 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0330 08:37:08.723472 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.574074 | |
| I0330 08:37:08.723489 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34028 (* 0.3 = 0.702084 loss) | |
| I0330 08:37:08.723505 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.738877 (* 0.3 = 0.221663 loss) | |
| I0330 08:37:08.723517 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.444444 | |
| I0330 08:37:08.723531 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 08:37:08.723551 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.722222 | |
| I0330 08:37:08.723564 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79101 (* 0.3 = 0.537302 loss) | |
| I0330 08:37:08.723578 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.585721 (* 0.3 = 0.175716 loss) | |
| I0330 08:37:08.723592 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.796296 | |
| I0330 08:37:08.723604 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 08:37:08.723616 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0330 08:37:08.723630 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.827576 (* 1 = 0.827576 loss) | |
| I0330 08:37:08.723646 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.291791 (* 1 = 0.291791 loss) | |
| I0330 08:37:08.723669 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 08:37:08.723691 10583 solver.cpp:245] Train net output #16: total_confidence = 0.194031 | |
| I0330 08:37:08.723707 10583 sgd_solver.cpp:106] Iteration 103500, lr = 0.01 | |
| I0330 08:39:18.200841 10583 solver.cpp:229] Iteration 104000, loss = 3.18608 | |
| I0330 08:39:18.200953 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.315789 | |
| I0330 08:39:18.200973 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 08:39:18.200985 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.605263 | |
| I0330 08:39:18.201002 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.37251 (* 0.3 = 0.711753 loss) | |
| I0330 08:39:18.201017 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.669927 (* 0.3 = 0.200978 loss) | |
| I0330 08:39:18.201030 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.394737 | |
| I0330 08:39:18.201041 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 08:39:18.201053 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.631579 | |
| I0330 08:39:18.201066 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.0944 (* 0.3 = 0.628321 loss) | |
| I0330 08:39:18.201082 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.551515 (* 0.3 = 0.165454 loss) | |
| I0330 08:39:18.201094 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.552632 | |
| I0330 08:39:18.201107 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 08:39:18.201118 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.736842 | |
| I0330 08:39:18.201133 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.58648 (* 1 = 1.58648 loss) | |
| I0330 08:39:18.201146 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.448931 (* 1 = 0.448931 loss) | |
| I0330 08:39:18.201160 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:39:18.201174 10583 solver.cpp:245] Train net output #16: total_confidence = 0.231113 | |
| I0330 08:39:18.201185 10583 sgd_solver.cpp:106] Iteration 104000, lr = 0.01 | |
| I0330 08:41:27.573396 10583 solver.cpp:229] Iteration 104500, loss = 3.20032 | |
| I0330 08:41:27.573539 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.28 | |
| I0330 08:41:27.573560 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 08:41:27.573582 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.52 | |
| I0330 08:41:27.573598 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.28767 (* 0.3 = 0.686301 loss) | |
| I0330 08:41:27.573613 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.773084 (* 0.3 = 0.231925 loss) | |
| I0330 08:41:27.573626 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.34 | |
| I0330 08:41:27.573639 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0330 08:41:27.573652 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.66 | |
| I0330 08:41:27.573665 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04821 (* 0.3 = 0.614464 loss) | |
| I0330 08:41:27.573679 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.667032 (* 0.3 = 0.200109 loss) | |
| I0330 08:41:27.573691 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0330 08:41:27.573704 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 08:41:27.573715 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0330 08:41:27.573729 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.11386 (* 1 = 1.11386 loss) | |
| I0330 08:41:27.573745 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.357007 (* 1 = 0.357007 loss) | |
| I0330 08:41:27.573756 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:41:27.573768 10583 solver.cpp:245] Train net output #16: total_confidence = 0.242393 | |
| I0330 08:41:27.573781 10583 sgd_solver.cpp:106] Iteration 104500, lr = 0.01 | |
| I0330 08:43:36.772205 10583 solver.cpp:338] Iteration 105000, Testing net (#0) | |
| I0330 08:44:06.606665 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 08:44:06.606717 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 08:44:06.606735 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760273 | |
| I0330 08:44:06.606748 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 08:44:06.606765 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:44:06.606781 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:44:06.606794 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 08:44:06.606806 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760273 | |
| I0330 08:44:06.606817 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 08:44:06.606835 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:44:06.606850 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 08:44:06.606863 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 08:44:06.606874 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760273 | |
| I0330 08:44:06.606895 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 08:44:06.606909 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 08:44:06.606923 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 08:44:06.606936 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 08:44:06.606948 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 08:44:06.758718 10583 solver.cpp:229] Iteration 105000, loss = 3.17128 | |
| I0330 08:44:06.758759 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 08:44:06.758774 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 08:44:06.758787 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.625 | |
| I0330 08:44:06.758802 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.05824 (* 0.3 = 0.617473 loss) | |
| I0330 08:44:06.758824 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.573157 (* 0.3 = 0.171947 loss) | |
| I0330 08:44:06.758841 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.475 | |
| I0330 08:44:06.758852 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 08:44:06.758865 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 08:44:06.758888 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61023 (* 0.3 = 0.483069 loss) | |
| I0330 08:44:06.758901 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.549988 (* 0.3 = 0.164996 loss) | |
| I0330 08:44:06.758914 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.725 | |
| I0330 08:44:06.758926 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 08:44:06.758937 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0330 08:44:06.758952 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14382 (* 1 = 1.14382 loss) | |
| I0330 08:44:06.758966 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.361645 (* 1 = 0.361645 loss) | |
| I0330 08:44:06.758996 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:44:06.759019 10583 solver.cpp:245] Train net output #16: total_confidence = 0.164632 | |
| I0330 08:44:06.759032 10583 sgd_solver.cpp:106] Iteration 105000, lr = 0.01 | |
| I0330 08:46:16.086462 10583 solver.cpp:229] Iteration 105500, loss = 3.19551 | |
| I0330 08:46:16.086616 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.304348 | |
| I0330 08:46:16.086637 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 08:46:16.086650 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 08:46:16.086668 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.55634 (* 0.3 = 0.766903 loss) | |
| I0330 08:46:16.086689 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.744898 (* 0.3 = 0.223469 loss) | |
| I0330 08:46:16.086702 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.434783 | |
| I0330 08:46:16.086714 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 08:46:16.086727 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.608696 | |
| I0330 08:46:16.086741 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.52165 (* 0.3 = 0.756494 loss) | |
| I0330 08:46:16.086755 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.719557 (* 0.3 = 0.215867 loss) | |
| I0330 08:46:16.086767 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 08:46:16.086781 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 08:46:16.086792 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.608696 | |
| I0330 08:46:16.086807 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.92772 (* 1 = 1.92772 loss) | |
| I0330 08:46:16.086828 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.554959 (* 1 = 0.554959 loss) | |
| I0330 08:46:16.086841 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:46:16.086853 10583 solver.cpp:245] Train net output #16: total_confidence = 0.241155 | |
| I0330 08:46:16.086865 10583 sgd_solver.cpp:106] Iteration 105500, lr = 0.01 | |
| I0330 08:48:25.422397 10583 solver.cpp:229] Iteration 106000, loss = 3.18244 | |
| I0330 08:48:25.422546 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 08:48:25.422566 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 08:48:25.422580 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 08:48:25.422596 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15668 (* 0.3 = 0.647005 loss) | |
| I0330 08:48:25.422621 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.623639 (* 0.3 = 0.187092 loss) | |
| I0330 08:48:25.422632 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.422222 | |
| I0330 08:48:25.422644 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 08:48:25.422657 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0330 08:48:25.422672 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.74538 (* 0.3 = 0.523615 loss) | |
| I0330 08:48:25.422685 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.516383 (* 0.3 = 0.154915 loss) | |
| I0330 08:48:25.422698 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.8 | |
| I0330 08:48:25.422711 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 08:48:25.422724 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0330 08:48:25.422739 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.669134 (* 1 = 0.669134 loss) | |
| I0330 08:48:25.422752 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.206818 (* 1 = 0.206818 loss) | |
| I0330 08:48:25.422765 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:48:25.422776 10583 solver.cpp:245] Train net output #16: total_confidence = 0.229655 | |
| I0330 08:48:25.422788 10583 sgd_solver.cpp:106] Iteration 106000, lr = 0.01 | |
| I0330 08:50:34.788161 10583 solver.cpp:229] Iteration 106500, loss = 3.19882 | |
| I0330 08:50:34.788295 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.325581 | |
| I0330 08:50:34.788316 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 08:50:34.788328 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.534884 | |
| I0330 08:50:34.788346 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.21574 (* 0.3 = 0.664723 loss) | |
| I0330 08:50:34.788362 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.62344 (* 0.3 = 0.187032 loss) | |
| I0330 08:50:34.788373 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488372 | |
| I0330 08:50:34.788386 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 08:50:34.788398 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.697674 | |
| I0330 08:50:34.788411 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6894 (* 0.3 = 0.506819 loss) | |
| I0330 08:50:34.788426 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.470191 (* 0.3 = 0.141057 loss) | |
| I0330 08:50:34.788439 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.697674 | |
| I0330 08:50:34.788450 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 08:50:34.788462 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0330 08:50:34.788476 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00045 (* 1 = 1.00045 loss) | |
| I0330 08:50:34.788491 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.287344 (* 1 = 0.287344 loss) | |
| I0330 08:50:34.788502 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 08:50:34.788519 10583 solver.cpp:245] Train net output #16: total_confidence = 0.101463 | |
| I0330 08:50:34.788532 10583 sgd_solver.cpp:106] Iteration 106500, lr = 0.01 | |
| I0330 08:52:44.110496 10583 solver.cpp:229] Iteration 107000, loss = 3.10464 | |
| I0330 08:52:44.110637 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0330 08:52:44.110658 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 08:52:44.110671 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0330 08:52:44.110688 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.40294 (* 0.3 = 0.720882 loss) | |
| I0330 08:52:44.110703 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.703779 (* 0.3 = 0.211134 loss) | |
| I0330 08:52:44.110720 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0330 08:52:44.110733 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 08:52:44.110745 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.520833 | |
| I0330 08:52:44.110759 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.11546 (* 0.3 = 0.634639 loss) | |
| I0330 08:52:44.110774 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.632822 (* 0.3 = 0.189847 loss) | |
| I0330 08:52:44.110786 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 08:52:44.110798 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 08:52:44.110810 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.645833 | |
| I0330 08:52:44.110829 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.5085 (* 1 = 1.5085 loss) | |
| I0330 08:52:44.110843 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.461527 (* 1 = 0.461527 loss) | |
| I0330 08:52:44.110855 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 08:52:44.110867 10583 solver.cpp:245] Train net output #16: total_confidence = 0.157892 | |
| I0330 08:52:44.110888 10583 sgd_solver.cpp:106] Iteration 107000, lr = 0.01 | |
| I0330 08:54:53.304906 10583 solver.cpp:229] Iteration 107500, loss = 3.16195 | |
| I0330 08:54:53.305018 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.42 | |
| I0330 08:54:53.305038 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 08:54:53.305052 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.66 | |
| I0330 08:54:53.305069 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14437 (* 0.3 = 0.643312 loss) | |
| I0330 08:54:53.305083 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.663132 (* 0.3 = 0.198939 loss) | |
| I0330 08:54:53.305096 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.46 | |
| I0330 08:54:53.305109 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 08:54:53.305120 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0330 08:54:53.305135 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79307 (* 0.3 = 0.537922 loss) | |
| I0330 08:54:53.305150 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524023 (* 0.3 = 0.157207 loss) | |
| I0330 08:54:53.305166 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.78 | |
| I0330 08:54:53.305181 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 08:54:53.305192 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0330 08:54:53.305207 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.788009 (* 1 = 0.788009 loss) | |
| I0330 08:54:53.305222 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.261557 (* 1 = 0.261557 loss) | |
| I0330 08:54:53.305233 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 08:54:53.305246 10583 solver.cpp:245] Train net output #16: total_confidence = 0.142496 | |
| I0330 08:54:53.305258 10583 sgd_solver.cpp:106] Iteration 107500, lr = 0.01 | |
| I0330 08:57:02.415948 10583 solver.cpp:229] Iteration 108000, loss = 3.1754 | |
| I0330 08:57:02.416049 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 08:57:02.416076 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 08:57:02.416090 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.657143 | |
| I0330 08:57:02.416106 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15216 (* 0.3 = 0.645648 loss) | |
| I0330 08:57:02.416121 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.665415 (* 0.3 = 0.199624 loss) | |
| I0330 08:57:02.416134 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.571429 | |
| I0330 08:57:02.416154 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 08:57:02.416167 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 08:57:02.416180 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.42168 (* 0.3 = 0.426504 loss) | |
| I0330 08:57:02.416194 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.460697 (* 0.3 = 0.138209 loss) | |
| I0330 08:57:02.416208 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0330 08:57:02.416219 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 08:57:02.416231 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.828571 | |
| I0330 08:57:02.416246 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.06022 (* 1 = 1.06022 loss) | |
| I0330 08:57:02.416260 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.337965 (* 1 = 0.337965 loss) | |
| I0330 08:57:02.416273 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 08:57:02.416285 10583 solver.cpp:245] Train net output #16: total_confidence = 0.239155 | |
| I0330 08:57:02.416297 10583 sgd_solver.cpp:106] Iteration 108000, lr = 0.01 | |
| I0330 08:59:11.552973 10583 solver.cpp:229] Iteration 108500, loss = 3.10645 | |
| I0330 08:59:11.553102 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.283019 | |
| I0330 08:59:11.553125 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 08:59:11.553138 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.471698 | |
| I0330 08:59:11.553155 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46967 (* 0.3 = 0.7409 loss) | |
| I0330 08:59:11.553174 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.787045 (* 0.3 = 0.236113 loss) | |
| I0330 08:59:11.553186 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.358491 | |
| I0330 08:59:11.553200 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 08:59:11.553211 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.641509 | |
| I0330 08:59:11.553225 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.14553 (* 0.3 = 0.64366 loss) | |
| I0330 08:59:11.553241 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.66103 (* 0.3 = 0.198309 loss) | |
| I0330 08:59:11.553252 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.603774 | |
| I0330 08:59:11.553266 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0330 08:59:11.553277 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.830189 | |
| I0330 08:59:11.553292 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.20406 (* 1 = 1.20406 loss) | |
| I0330 08:59:11.553305 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.367619 (* 1 = 0.367619 loss) | |
| I0330 08:59:11.553318 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 08:59:11.553329 10583 solver.cpp:245] Train net output #16: total_confidence = 0.119716 | |
| I0330 08:59:11.553341 10583 sgd_solver.cpp:106] Iteration 108500, lr = 0.01 | |
| I0330 09:01:20.642066 10583 solver.cpp:229] Iteration 109000, loss = 3.11679 | |
| I0330 09:01:20.642206 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3125 | |
| I0330 09:01:20.642232 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 09:01:20.642252 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.708333 | |
| I0330 09:01:20.642269 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.03811 (* 0.3 = 0.611434 loss) | |
| I0330 09:01:20.642284 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.619931 (* 0.3 = 0.185979 loss) | |
| I0330 09:01:20.642297 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.479167 | |
| I0330 09:01:20.642310 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 09:01:20.642323 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 09:01:20.642338 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7824 (* 0.3 = 0.534721 loss) | |
| I0330 09:01:20.642352 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.545179 (* 0.3 = 0.163554 loss) | |
| I0330 09:01:20.642364 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.770833 | |
| I0330 09:01:20.642377 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 09:01:20.642390 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.854167 | |
| I0330 09:01:20.642403 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.798555 (* 1 = 0.798555 loss) | |
| I0330 09:01:20.642417 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.233076 (* 1 = 0.233076 loss) | |
| I0330 09:01:20.642429 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:01:20.642441 10583 solver.cpp:245] Train net output #16: total_confidence = 0.209955 | |
| I0330 09:01:20.642453 10583 sgd_solver.cpp:106] Iteration 109000, lr = 0.01 | |
| I0330 09:03:29.892419 10583 solver.cpp:229] Iteration 109500, loss = 3.1561 | |
| I0330 09:03:29.892545 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 09:03:29.892566 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 09:03:29.892580 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0330 09:03:29.892596 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.43228 (* 0.3 = 0.729685 loss) | |
| I0330 09:03:29.892611 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.716136 (* 0.3 = 0.214841 loss) | |
| I0330 09:03:29.892624 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.354167 | |
| I0330 09:03:29.892637 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 09:03:29.892648 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.541667 | |
| I0330 09:03:29.892663 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.21538 (* 0.3 = 0.664614 loss) | |
| I0330 09:03:29.892676 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.645338 (* 0.3 = 0.193602 loss) | |
| I0330 09:03:29.892689 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.541667 | |
| I0330 09:03:29.892701 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 09:03:29.892714 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0330 09:03:29.892727 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60827 (* 1 = 1.60827 loss) | |
| I0330 09:03:29.892741 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.448176 (* 1 = 0.448176 loss) | |
| I0330 09:03:29.892753 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:03:29.892766 10583 solver.cpp:245] Train net output #16: total_confidence = 0.119586 | |
| I0330 09:03:29.892779 10583 sgd_solver.cpp:106] Iteration 109500, lr = 0.01 | |
| I0330 09:05:39.599094 10583 solver.cpp:338] Iteration 110000, Testing net (#0) | |
| I0330 09:06:09.474030 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 09:06:09.474092 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 09:06:09.474108 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.759274 | |
| I0330 09:06:09.474123 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 09:06:09.474139 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:06:09.474156 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:06:09.474171 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 09:06:09.474184 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.759274 | |
| I0330 09:06:09.474196 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 09:06:09.474210 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:06:09.474225 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:06:09.474237 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 09:06:09.474249 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.759274 | |
| I0330 09:06:09.474261 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 09:06:09.474275 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:06:09.474289 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:06:09.474301 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 09:06:09.474318 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 09:06:09.626394 10583 solver.cpp:229] Iteration 110000, loss = 3.06199 | |
| I0330 09:06:09.626534 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.357143 | |
| I0330 09:06:09.626555 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 09:06:09.626569 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 09:06:09.626586 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.26918 (* 0.3 = 0.680753 loss) | |
| I0330 09:06:09.626601 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.642029 (* 0.3 = 0.192609 loss) | |
| I0330 09:06:09.626613 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 09:06:09.626626 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 09:06:09.626638 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0330 09:06:09.626653 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.71184 (* 0.3 = 0.513551 loss) | |
| I0330 09:06:09.626668 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.452888 (* 0.3 = 0.135866 loss) | |
| I0330 09:06:09.626680 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.619048 | |
| I0330 09:06:09.626693 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 09:06:09.626704 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 09:06:09.626718 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.21594 (* 1 = 1.21594 loss) | |
| I0330 09:06:09.626732 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.306814 (* 1 = 0.306814 loss) | |
| I0330 09:06:09.626744 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:06:09.626756 10583 solver.cpp:245] Train net output #16: total_confidence = 0.202197 | |
| I0330 09:06:09.626770 10583 sgd_solver.cpp:106] Iteration 110000, lr = 0.01 | |
| I0330 09:08:18.717779 10583 solver.cpp:229] Iteration 110500, loss = 3.09077 | |
| I0330 09:08:18.717964 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.325 | |
| I0330 09:08:18.717985 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 09:08:18.717999 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55 | |
| I0330 09:08:18.718015 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.31764 (* 0.3 = 0.695291 loss) | |
| I0330 09:08:18.718031 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.731732 (* 0.3 = 0.21952 loss) | |
| I0330 09:08:18.718044 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45 | |
| I0330 09:08:18.718056 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 09:08:18.718068 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.625 | |
| I0330 09:08:18.718083 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.00482 (* 0.3 = 0.601446 loss) | |
| I0330 09:08:18.718097 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.584621 (* 0.3 = 0.175386 loss) | |
| I0330 09:08:18.718111 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 09:08:18.718122 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 09:08:18.718135 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 09:08:18.718149 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.992619 (* 1 = 0.992619 loss) | |
| I0330 09:08:18.718166 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.333412 (* 1 = 0.333412 loss) | |
| I0330 09:08:18.718179 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:08:18.718191 10583 solver.cpp:245] Train net output #16: total_confidence = 0.202138 | |
| I0330 09:08:18.718204 10583 sgd_solver.cpp:106] Iteration 110500, lr = 0.01 | |
| I0330 09:10:28.084630 10583 solver.cpp:229] Iteration 111000, loss = 3.06884 | |
| I0330 09:10:28.084748 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395833 | |
| I0330 09:10:28.084767 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 09:10:28.084780 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.708333 | |
| I0330 09:10:28.084796 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.99405 (* 0.3 = 0.598216 loss) | |
| I0330 09:10:28.084811 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.58494 (* 0.3 = 0.175482 loss) | |
| I0330 09:10:28.084825 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.458333 | |
| I0330 09:10:28.084837 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 09:10:28.084849 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.604167 | |
| I0330 09:10:28.084863 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6702 (* 0.3 = 0.50106 loss) | |
| I0330 09:10:28.084877 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.495938 (* 0.3 = 0.148781 loss) | |
| I0330 09:10:28.084889 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0330 09:10:28.084902 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 09:10:28.084913 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0330 09:10:28.084928 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.765167 (* 1 = 0.765167 loss) | |
| I0330 09:10:28.084941 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.232425 (* 1 = 0.232425 loss) | |
| I0330 09:10:28.084954 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:10:28.084965 10583 solver.cpp:245] Train net output #16: total_confidence = 0.250953 | |
| I0330 09:10:28.084977 10583 sgd_solver.cpp:106] Iteration 111000, lr = 0.01 | |
| I0330 09:12:37.240670 10583 solver.cpp:229] Iteration 111500, loss = 3.09267 | |
| I0330 09:12:37.240808 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.355556 | |
| I0330 09:12:37.240828 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 09:12:37.240841 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 09:12:37.240859 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.05092 (* 0.3 = 0.615277 loss) | |
| I0330 09:12:37.240874 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.54653 (* 0.3 = 0.163959 loss) | |
| I0330 09:12:37.240885 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.355556 | |
| I0330 09:12:37.240898 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 09:12:37.240911 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 09:12:37.240924 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.73055 (* 0.3 = 0.519166 loss) | |
| I0330 09:12:37.240939 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.467844 (* 0.3 = 0.140353 loss) | |
| I0330 09:12:37.240952 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.733333 | |
| I0330 09:12:37.240964 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 09:12:37.240977 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.866667 | |
| I0330 09:12:37.240991 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.842728 (* 1 = 0.842728 loss) | |
| I0330 09:12:37.241005 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.231753 (* 1 = 0.231753 loss) | |
| I0330 09:12:37.241019 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:12:37.241030 10583 solver.cpp:245] Train net output #16: total_confidence = 0.160575 | |
| I0330 09:12:37.241044 10583 sgd_solver.cpp:106] Iteration 111500, lr = 0.01 | |
| I0330 09:14:46.565208 10583 solver.cpp:229] Iteration 112000, loss = 3.10044 | |
| I0330 09:14:46.565318 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 09:14:46.565338 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 09:14:46.565351 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.642857 | |
| I0330 09:14:46.565367 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.45063 (* 0.3 = 0.73519 loss) | |
| I0330 09:14:46.565382 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.70667 (* 0.3 = 0.212001 loss) | |
| I0330 09:14:46.565395 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.47619 | |
| I0330 09:14:46.565407 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 09:14:46.565419 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.761905 | |
| I0330 09:14:46.565433 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64013 (* 0.3 = 0.49204 loss) | |
| I0330 09:14:46.565448 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.471706 (* 0.3 = 0.141512 loss) | |
| I0330 09:14:46.565460 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0330 09:14:46.565472 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 09:14:46.565485 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0330 09:14:46.565500 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.863123 (* 1 = 0.863123 loss) | |
| I0330 09:14:46.565513 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.318927 (* 1 = 0.318927 loss) | |
| I0330 09:14:46.565526 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:14:46.565537 10583 solver.cpp:245] Train net output #16: total_confidence = 0.116121 | |
| I0330 09:14:46.565549 10583 sgd_solver.cpp:106] Iteration 112000, lr = 0.01 | |
| I0330 09:16:55.914007 10583 solver.cpp:229] Iteration 112500, loss = 3.08579 | |
| I0330 09:16:55.914160 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0330 09:16:55.914183 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 09:16:55.914197 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.479167 | |
| I0330 09:16:55.914213 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.60642 (* 0.3 = 0.781926 loss) | |
| I0330 09:16:55.914235 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.761364 (* 0.3 = 0.228409 loss) | |
| I0330 09:16:55.914248 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4375 | |
| I0330 09:16:55.914261 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 09:16:55.914273 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.625 | |
| I0330 09:16:55.914286 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.98639 (* 0.3 = 0.595917 loss) | |
| I0330 09:16:55.914300 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.583392 (* 0.3 = 0.175018 loss) | |
| I0330 09:16:55.914312 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.645833 | |
| I0330 09:16:55.914324 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 09:16:55.914336 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 09:16:55.914351 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.32489 (* 1 = 1.32489 loss) | |
| I0330 09:16:55.914366 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.385504 (* 1 = 0.385504 loss) | |
| I0330 09:16:55.914377 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:16:55.914388 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0827415 | |
| I0330 09:16:55.914402 10583 sgd_solver.cpp:106] Iteration 112500, lr = 0.01 | |
| I0330 09:19:05.227890 10583 solver.cpp:229] Iteration 113000, loss = 3.09134 | |
| I0330 09:19:05.228008 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26 | |
| I0330 09:19:05.228027 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 09:19:05.228040 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0330 09:19:05.228056 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23952 (* 0.3 = 0.671855 loss) | |
| I0330 09:19:05.228071 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650401 (* 0.3 = 0.19512 loss) | |
| I0330 09:19:05.228083 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.38 | |
| I0330 09:19:05.228096 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 09:19:05.228107 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76 | |
| I0330 09:19:05.228121 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77062 (* 0.3 = 0.531185 loss) | |
| I0330 09:19:05.228135 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.516391 (* 0.3 = 0.154917 loss) | |
| I0330 09:19:05.228148 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.88 | |
| I0330 09:19:05.228163 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0330 09:19:05.228175 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.96 | |
| I0330 09:19:05.228189 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.535191 (* 1 = 0.535191 loss) | |
| I0330 09:19:05.228204 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.156693 (* 1 = 0.156693 loss) | |
| I0330 09:19:05.228215 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 09:19:05.228229 10583 solver.cpp:245] Train net output #16: total_confidence = 0.30219 | |
| I0330 09:19:05.228240 10583 sgd_solver.cpp:106] Iteration 113000, lr = 0.01 | |
| I0330 09:21:14.289685 10583 solver.cpp:229] Iteration 113500, loss = 3.04782 | |
| I0330 09:21:14.289834 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.317073 | |
| I0330 09:21:14.289854 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 09:21:14.289867 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.536585 | |
| I0330 09:21:14.289883 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.40461 (* 0.3 = 0.721383 loss) | |
| I0330 09:21:14.289898 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.685581 (* 0.3 = 0.205674 loss) | |
| I0330 09:21:14.289917 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.414634 | |
| I0330 09:21:14.289929 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 09:21:14.289942 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.682927 | |
| I0330 09:21:14.289955 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.89108 (* 0.3 = 0.567325 loss) | |
| I0330 09:21:14.289970 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.579328 (* 0.3 = 0.173799 loss) | |
| I0330 09:21:14.289983 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829268 | |
| I0330 09:21:14.289995 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 09:21:14.290007 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.951219 | |
| I0330 09:21:14.290021 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.604411 (* 1 = 0.604411 loss) | |
| I0330 09:21:14.290035 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.182233 (* 1 = 0.182233 loss) | |
| I0330 09:21:14.290047 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:21:14.290060 10583 solver.cpp:245] Train net output #16: total_confidence = 0.228998 | |
| I0330 09:21:14.290071 10583 sgd_solver.cpp:106] Iteration 113500, lr = 0.01 | |
| I0330 09:23:23.473073 10583 solver.cpp:229] Iteration 114000, loss = 3.09224 | |
| I0330 09:23:23.473201 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.469388 | |
| I0330 09:23:23.473220 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 09:23:23.473234 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.734694 | |
| I0330 09:23:23.473250 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78547 (* 0.3 = 0.535641 loss) | |
| I0330 09:23:23.473265 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.51205 (* 0.3 = 0.153615 loss) | |
| I0330 09:23:23.473278 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0330 09:23:23.473291 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 09:23:23.473304 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.836735 | |
| I0330 09:23:23.473317 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.504 (* 0.3 = 0.451199 loss) | |
| I0330 09:23:23.473333 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.432995 (* 0.3 = 0.129898 loss) | |
| I0330 09:23:23.473345 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.734694 | |
| I0330 09:23:23.473357 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 09:23:23.473369 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.918367 | |
| I0330 09:23:23.473383 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.842231 (* 1 = 0.842231 loss) | |
| I0330 09:23:23.473397 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.248041 (* 1 = 0.248041 loss) | |
| I0330 09:23:23.473409 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:23:23.473423 10583 solver.cpp:245] Train net output #16: total_confidence = 0.20752 | |
| I0330 09:23:23.473434 10583 sgd_solver.cpp:106] Iteration 114000, lr = 0.01 | |
| I0330 09:25:32.456553 10583 solver.cpp:229] Iteration 114500, loss = 3.07169 | |
| I0330 09:25:32.456701 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26 | |
| I0330 09:25:32.456720 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 09:25:32.456737 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 09:25:32.456771 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.26568 (* 0.3 = 0.679703 loss) | |
| I0330 09:25:32.456789 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.714714 (* 0.3 = 0.214414 loss) | |
| I0330 09:25:32.456801 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0330 09:25:32.456815 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 09:25:32.456826 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.68 | |
| I0330 09:25:32.456840 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.85573 (* 0.3 = 0.556719 loss) | |
| I0330 09:25:32.456856 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.59215 (* 0.3 = 0.177645 loss) | |
| I0330 09:25:32.456868 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0330 09:25:32.456881 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 09:25:32.456892 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0330 09:25:32.456905 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.938808 (* 1 = 0.938808 loss) | |
| I0330 09:25:32.456920 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.328793 (* 1 = 0.328793 loss) | |
| I0330 09:25:32.456933 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:25:32.456944 10583 solver.cpp:245] Train net output #16: total_confidence = 0.141636 | |
| I0330 09:25:32.456956 10583 sgd_solver.cpp:106] Iteration 114500, lr = 0.01 | |
| I0330 09:27:41.511684 10583 solver.cpp:338] Iteration 115000, Testing net (#0) | |
| I0330 09:28:11.340679 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 09:28:11.340739 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 09:28:11.340756 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.760092 | |
| I0330 09:28:11.340770 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 09:28:11.340785 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:28:11.340801 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:28:11.340813 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 09:28:11.340826 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.760092 | |
| I0330 09:28:11.340837 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 09:28:11.340850 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:28:11.340865 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:28:11.340878 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 09:28:11.340889 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.760092 | |
| I0330 09:28:11.340901 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 09:28:11.340915 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:28:11.340930 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:28:11.340942 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 09:28:11.340955 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 09:28:11.492458 10583 solver.cpp:229] Iteration 115000, loss = 3.0479 | |
| I0330 09:28:11.492501 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0330 09:28:11.492517 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 09:28:11.492530 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.738095 | |
| I0330 09:28:11.492545 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78325 (* 0.3 = 0.534974 loss) | |
| I0330 09:28:11.492560 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.528669 (* 0.3 = 0.158601 loss) | |
| I0330 09:28:11.492573 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.619048 | |
| I0330 09:28:11.492585 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 09:28:11.492597 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0330 09:28:11.492611 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.35964 (* 0.3 = 0.407891 loss) | |
| I0330 09:28:11.492625 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.409359 (* 0.3 = 0.122808 loss) | |
| I0330 09:28:11.492640 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.952381 | |
| I0330 09:28:11.492655 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.988636 | |
| I0330 09:28:11.492666 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 09:28:11.492681 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.232895 (* 1 = 0.232895 loss) | |
| I0330 09:28:11.492696 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0614596 (* 1 = 0.0614596 loss) | |
| I0330 09:28:11.492707 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 09:28:11.492719 10583 solver.cpp:245] Train net output #16: total_confidence = 0.430854 | |
| I0330 09:28:11.492732 10583 sgd_solver.cpp:106] Iteration 115000, lr = 0.01 | |
| I0330 09:30:20.733281 10583 solver.cpp:229] Iteration 115500, loss = 3.05176 | |
| I0330 09:30:20.733436 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27907 | |
| I0330 09:30:20.733458 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 09:30:20.733470 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.488372 | |
| I0330 09:30:20.733486 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.6024 (* 0.3 = 0.78072 loss) | |
| I0330 09:30:20.733506 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.747735 (* 0.3 = 0.22432 loss) | |
| I0330 09:30:20.733518 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.395349 | |
| I0330 09:30:20.733530 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 09:30:20.733542 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.55814 | |
| I0330 09:30:20.733556 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.0441 (* 0.3 = 0.613231 loss) | |
| I0330 09:30:20.733569 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.584776 (* 0.3 = 0.175433 loss) | |
| I0330 09:30:20.733582 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.581395 | |
| I0330 09:30:20.733594 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 09:30:20.733606 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.790698 | |
| I0330 09:30:20.733620 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.40903 (* 1 = 1.40903 loss) | |
| I0330 09:30:20.733633 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.408039 (* 1 = 0.408039 loss) | |
| I0330 09:30:20.733646 10583 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 09:30:20.733657 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0552123 | |
| I0330 09:30:20.733670 10583 sgd_solver.cpp:106] Iteration 115500, lr = 0.01 | |
| I0330 09:32:29.970250 10583 solver.cpp:229] Iteration 116000, loss = 3.01523 | |
| I0330 09:32:29.970403 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 09:32:29.970425 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 09:32:29.970438 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.517857 | |
| I0330 09:32:29.970454 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4871 (* 0.3 = 0.746129 loss) | |
| I0330 09:32:29.970469 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.814844 (* 0.3 = 0.244453 loss) | |
| I0330 09:32:29.970489 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.428571 | |
| I0330 09:32:29.970500 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 09:32:29.970513 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.642857 | |
| I0330 09:32:29.970527 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96447 (* 0.3 = 0.58934 loss) | |
| I0330 09:32:29.970541 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.634068 (* 0.3 = 0.19022 loss) | |
| I0330 09:32:29.970554 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.678571 | |
| I0330 09:32:29.970566 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 09:32:29.970577 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0330 09:32:29.970592 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.98134 (* 1 = 0.98134 loss) | |
| I0330 09:32:29.970607 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.331903 (* 1 = 0.331903 loss) | |
| I0330 09:32:29.970618 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:32:29.970630 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0927533 | |
| I0330 09:32:29.970643 10583 sgd_solver.cpp:106] Iteration 116000, lr = 0.01 | |
| I0330 09:34:39.430100 10583 solver.cpp:229] Iteration 116500, loss = 3.06749 | |
| I0330 09:34:39.430215 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0330 09:34:39.430236 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 09:34:39.430249 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.625 | |
| I0330 09:34:39.430265 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23904 (* 0.3 = 0.671712 loss) | |
| I0330 09:34:39.430280 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.626412 (* 0.3 = 0.187924 loss) | |
| I0330 09:34:39.430292 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.541667 | |
| I0330 09:34:39.430305 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 09:34:39.430316 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 09:34:39.430330 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9365 (* 0.3 = 0.580951 loss) | |
| I0330 09:34:39.430346 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.558291 (* 0.3 = 0.167487 loss) | |
| I0330 09:34:39.430358 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.645833 | |
| I0330 09:34:39.430371 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 09:34:39.430382 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 09:34:39.430397 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28156 (* 1 = 1.28156 loss) | |
| I0330 09:34:39.430409 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.355532 (* 1 = 0.355532 loss) | |
| I0330 09:34:39.430423 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:34:39.430434 10583 solver.cpp:245] Train net output #16: total_confidence = 0.276429 | |
| I0330 09:34:39.430446 10583 sgd_solver.cpp:106] Iteration 116500, lr = 0.01 | |
| I0330 09:36:48.606251 10583 solver.cpp:229] Iteration 117000, loss = 3.09047 | |
| I0330 09:36:48.606411 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.214286 | |
| I0330 09:36:48.606432 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 09:36:48.606446 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.446429 | |
| I0330 09:36:48.606462 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.66835 (* 0.3 = 0.800505 loss) | |
| I0330 09:36:48.606477 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.855351 (* 0.3 = 0.256605 loss) | |
| I0330 09:36:48.606498 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.428571 | |
| I0330 09:36:48.606509 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 09:36:48.606523 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.696429 | |
| I0330 09:36:48.606536 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.18124 (* 0.3 = 0.654372 loss) | |
| I0330 09:36:48.606550 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.702663 (* 0.3 = 0.210799 loss) | |
| I0330 09:36:48.606562 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0330 09:36:48.606575 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 09:36:48.606586 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0330 09:36:48.606600 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24425 (* 1 = 1.24425 loss) | |
| I0330 09:36:48.606614 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.400107 (* 1 = 0.400107 loss) | |
| I0330 09:36:48.606626 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:36:48.606638 10583 solver.cpp:245] Train net output #16: total_confidence = 0.113788 | |
| I0330 09:36:48.606650 10583 sgd_solver.cpp:106] Iteration 117000, lr = 0.01 | |
| I0330 09:38:57.891007 10583 solver.cpp:229] Iteration 117500, loss = 3.03554 | |
| I0330 09:38:57.891124 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27451 | |
| I0330 09:38:57.891144 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 09:38:57.891157 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.647059 | |
| I0330 09:38:57.891176 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09906 (* 0.3 = 0.629719 loss) | |
| I0330 09:38:57.891191 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.632852 (* 0.3 = 0.189856 loss) | |
| I0330 09:38:57.891204 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.490196 | |
| I0330 09:38:57.891216 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 09:38:57.891228 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.803922 | |
| I0330 09:38:57.891242 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64913 (* 0.3 = 0.494738 loss) | |
| I0330 09:38:57.891257 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.514363 (* 0.3 = 0.154309 loss) | |
| I0330 09:38:57.891268 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.745098 | |
| I0330 09:38:57.891280 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 09:38:57.891293 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.941176 | |
| I0330 09:38:57.891306 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.817325 (* 1 = 0.817325 loss) | |
| I0330 09:38:57.891320 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.243006 (* 1 = 0.243006 loss) | |
| I0330 09:38:57.891332 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:38:57.891345 10583 solver.cpp:245] Train net output #16: total_confidence = 0.203186 | |
| I0330 09:38:57.891356 10583 sgd_solver.cpp:106] Iteration 117500, lr = 0.01 | |
| I0330 09:41:07.436146 10583 solver.cpp:229] Iteration 118000, loss = 3.07011 | |
| I0330 09:41:07.436288 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.45 | |
| I0330 09:41:07.436307 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 09:41:07.436321 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.7 | |
| I0330 09:41:07.436337 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93295 (* 0.3 = 0.579885 loss) | |
| I0330 09:41:07.436359 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.543286 (* 0.3 = 0.162986 loss) | |
| I0330 09:41:07.436372 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.525 | |
| I0330 09:41:07.436384 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 09:41:07.436396 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.925 | |
| I0330 09:41:07.436411 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32677 (* 0.3 = 0.398032 loss) | |
| I0330 09:41:07.436425 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.384032 (* 0.3 = 0.11521 loss) | |
| I0330 09:41:07.436437 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.775 | |
| I0330 09:41:07.436450 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 09:41:07.436461 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 09:41:07.436476 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.853882 (* 1 = 0.853882 loss) | |
| I0330 09:41:07.436491 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.225346 (* 1 = 0.225346 loss) | |
| I0330 09:41:07.436502 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 09:41:07.436514 10583 solver.cpp:245] Train net output #16: total_confidence = 0.282959 | |
| I0330 09:41:07.436527 10583 sgd_solver.cpp:106] Iteration 118000, lr = 0.01 | |
| I0330 09:43:16.892427 10583 solver.cpp:229] Iteration 118500, loss = 3.05032 | |
| I0330 09:43:16.892554 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26 | |
| I0330 09:43:16.892575 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 09:43:16.892588 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0330 09:43:16.892604 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.55266 (* 0.3 = 0.765797 loss) | |
| I0330 09:43:16.892619 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.793027 (* 0.3 = 0.237908 loss) | |
| I0330 09:43:16.892632 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.38 | |
| I0330 09:43:16.892644 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 09:43:16.892657 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0330 09:43:16.892670 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06745 (* 0.3 = 0.620235 loss) | |
| I0330 09:43:16.892684 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.659677 (* 0.3 = 0.197903 loss) | |
| I0330 09:43:16.892696 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.62 | |
| I0330 09:43:16.892709 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 09:43:16.892719 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 09:43:16.892734 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28199 (* 1 = 1.28199 loss) | |
| I0330 09:43:16.892747 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.385222 (* 1 = 0.385222 loss) | |
| I0330 09:43:16.892760 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:43:16.892771 10583 solver.cpp:245] Train net output #16: total_confidence = 0.183663 | |
| I0330 09:43:16.892783 10583 sgd_solver.cpp:106] Iteration 118500, lr = 0.01 | |
| I0330 09:45:26.393123 10583 solver.cpp:229] Iteration 119000, loss = 3.00975 | |
| I0330 09:45:26.393301 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 09:45:26.393322 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 09:45:26.393337 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 09:45:26.393352 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.08001 (* 0.3 = 0.624002 loss) | |
| I0330 09:45:26.393376 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.64753 (* 0.3 = 0.194259 loss) | |
| I0330 09:45:26.393389 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.411765 | |
| I0330 09:45:26.393404 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 09:45:26.393415 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.764706 | |
| I0330 09:45:26.393429 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6916 (* 0.3 = 0.507479 loss) | |
| I0330 09:45:26.393443 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524907 (* 0.3 = 0.157472 loss) | |
| I0330 09:45:26.393455 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.686275 | |
| I0330 09:45:26.393467 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 09:45:26.393479 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.941176 | |
| I0330 09:45:26.393494 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.882072 (* 1 = 0.882072 loss) | |
| I0330 09:45:26.393508 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.281672 (* 1 = 0.281672 loss) | |
| I0330 09:45:26.393520 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:45:26.393534 10583 solver.cpp:245] Train net output #16: total_confidence = 0.127793 | |
| I0330 09:45:26.393556 10583 sgd_solver.cpp:106] Iteration 119000, lr = 0.01 | |
| I0330 09:47:35.928122 10583 solver.cpp:229] Iteration 119500, loss = 3.05531 | |
| I0330 09:47:35.928236 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.309524 | |
| I0330 09:47:35.928256 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 09:47:35.928270 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 09:47:35.928287 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.1954 (* 0.3 = 0.658619 loss) | |
| I0330 09:47:35.928302 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.631517 (* 0.3 = 0.189455 loss) | |
| I0330 09:47:35.928314 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.47619 | |
| I0330 09:47:35.928328 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 09:47:35.928339 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.880952 | |
| I0330 09:47:35.928352 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.46929 (* 0.3 = 0.440788 loss) | |
| I0330 09:47:35.928367 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.453065 (* 0.3 = 0.135919 loss) | |
| I0330 09:47:35.928380 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0330 09:47:35.928392 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 09:47:35.928405 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.952381 | |
| I0330 09:47:35.928418 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.697936 (* 1 = 0.697936 loss) | |
| I0330 09:47:35.928432 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.20891 (* 1 = 0.20891 loss) | |
| I0330 09:47:35.928444 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 09:47:35.928457 10583 solver.cpp:245] Train net output #16: total_confidence = 0.24651 | |
| I0330 09:47:35.928468 10583 sgd_solver.cpp:106] Iteration 119500, lr = 0.01 | |
| I0330 09:49:45.187692 10583 solver.cpp:338] Iteration 120000, Testing net (#0) | |
| I0330 09:50:15.008587 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 09:50:15.008638 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 09:50:15.008656 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.76041 | |
| I0330 09:50:15.008669 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 09:50:15.008687 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:50:15.008702 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:50:15.008715 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 09:50:15.008728 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.76041 | |
| I0330 09:50:15.008739 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 09:50:15.008754 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:50:15.008769 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 09:50:15.008780 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 09:50:15.008792 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.76041 | |
| I0330 09:50:15.008805 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 09:50:15.008818 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:50:15.008832 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 09:50:15.008846 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 09:50:15.008857 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 09:50:15.160099 10583 solver.cpp:229] Iteration 120000, loss = 3.01566 | |
| I0330 09:50:15.160140 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0330 09:50:15.160158 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 09:50:15.160172 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.681818 | |
| I0330 09:50:15.160187 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93392 (* 0.3 = 0.580177 loss) | |
| I0330 09:50:15.160202 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.565943 (* 0.3 = 0.169783 loss) | |
| I0330 09:50:15.160215 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.568182 | |
| I0330 09:50:15.160228 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 09:50:15.160240 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0330 09:50:15.160254 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.45075 (* 0.3 = 0.435226 loss) | |
| I0330 09:50:15.160269 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.431801 (* 0.3 = 0.12954 loss) | |
| I0330 09:50:15.160281 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.840909 | |
| I0330 09:50:15.160293 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 09:50:15.160305 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0330 09:50:15.160320 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.556174 (* 1 = 0.556174 loss) | |
| I0330 09:50:15.160333 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.144962 (* 1 = 0.144962 loss) | |
| I0330 09:50:15.160346 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:50:15.160358 10583 solver.cpp:245] Train net output #16: total_confidence = 0.400315 | |
| I0330 09:50:15.160372 10583 sgd_solver.cpp:106] Iteration 120000, lr = 0.01 | |
| I0330 09:52:24.585458 10583 solver.cpp:229] Iteration 120500, loss = 3.01348 | |
| I0330 09:52:24.585610 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340909 | |
| I0330 09:52:24.585630 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 09:52:24.585644 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.545455 | |
| I0330 09:52:24.585664 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39048 (* 0.3 = 0.717144 loss) | |
| I0330 09:52:24.585680 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.734691 (* 0.3 = 0.220407 loss) | |
| I0330 09:52:24.585692 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.409091 | |
| I0330 09:52:24.585705 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 09:52:24.585717 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.613636 | |
| I0330 09:52:24.585731 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37019 (* 0.3 = 0.711056 loss) | |
| I0330 09:52:24.585746 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.720942 (* 0.3 = 0.216283 loss) | |
| I0330 09:52:24.585757 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.545455 | |
| I0330 09:52:24.585769 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 09:52:24.585782 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.772727 | |
| I0330 09:52:24.585795 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.85835 (* 1 = 1.85835 loss) | |
| I0330 09:52:24.585809 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.534947 (* 1 = 0.534947 loss) | |
| I0330 09:52:24.585821 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:52:24.585834 10583 solver.cpp:245] Train net output #16: total_confidence = 0.264428 | |
| I0330 09:52:24.585845 10583 sgd_solver.cpp:106] Iteration 120500, lr = 0.01 | |
| I0330 09:54:34.625404 10583 solver.cpp:229] Iteration 121000, loss = 3.06463 | |
| I0330 09:54:34.625543 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.204082 | |
| I0330 09:54:34.625563 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0330 09:54:34.625578 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510204 | |
| I0330 09:54:34.625594 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41554 (* 0.3 = 0.724661 loss) | |
| I0330 09:54:34.625609 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.721015 (* 0.3 = 0.216304 loss) | |
| I0330 09:54:34.625622 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.530612 | |
| I0330 09:54:34.625634 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 09:54:34.625646 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.693878 | |
| I0330 09:54:34.625660 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75365 (* 0.3 = 0.526094 loss) | |
| I0330 09:54:34.625675 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.557736 (* 0.3 = 0.167321 loss) | |
| I0330 09:54:34.625687 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.632653 | |
| I0330 09:54:34.625699 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 09:54:34.625711 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0330 09:54:34.625726 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.21114 (* 1 = 1.21114 loss) | |
| I0330 09:54:34.625741 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.364353 (* 1 = 0.364353 loss) | |
| I0330 09:54:34.625752 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 09:54:34.625764 10583 solver.cpp:245] Train net output #16: total_confidence = 0.215151 | |
| I0330 09:54:34.625777 10583 sgd_solver.cpp:106] Iteration 121000, lr = 0.01 | |
| I0330 09:56:44.012902 10583 solver.cpp:229] Iteration 121500, loss = 2.98237 | |
| I0330 09:56:44.013051 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.415094 | |
| I0330 09:56:44.013070 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 09:56:44.013083 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.584906 | |
| I0330 09:56:44.013100 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17342 (* 0.3 = 0.652026 loss) | |
| I0330 09:56:44.013115 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.681728 (* 0.3 = 0.204518 loss) | |
| I0330 09:56:44.013135 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.415094 | |
| I0330 09:56:44.013149 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 09:56:44.013162 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.679245 | |
| I0330 09:56:44.013177 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88536 (* 0.3 = 0.565607 loss) | |
| I0330 09:56:44.013191 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.592134 (* 0.3 = 0.17764 loss) | |
| I0330 09:56:44.013203 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.660377 | |
| I0330 09:56:44.013216 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 09:56:44.013228 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.811321 | |
| I0330 09:56:44.013242 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24219 (* 1 = 1.24219 loss) | |
| I0330 09:56:44.013257 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.384674 (* 1 = 0.384674 loss) | |
| I0330 09:56:44.013269 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 09:56:44.013281 10583 solver.cpp:245] Train net output #16: total_confidence = 0.178105 | |
| I0330 09:56:44.013293 10583 sgd_solver.cpp:106] Iteration 121500, lr = 0.01 | |
| I0330 09:58:53.462693 10583 solver.cpp:229] Iteration 122000, loss = 3.00297 | |
| I0330 09:58:53.462810 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.369565 | |
| I0330 09:58:53.462829 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 09:58:53.462843 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.608696 | |
| I0330 09:58:53.462859 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.19967 (* 0.3 = 0.659902 loss) | |
| I0330 09:58:53.462874 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.658024 (* 0.3 = 0.197407 loss) | |
| I0330 09:58:53.462888 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.456522 | |
| I0330 09:58:53.462899 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 09:58:53.462920 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0330 09:58:53.462935 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.92748 (* 0.3 = 0.578245 loss) | |
| I0330 09:58:53.462949 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.562875 (* 0.3 = 0.168862 loss) | |
| I0330 09:58:53.462961 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.652174 | |
| I0330 09:58:53.462990 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 09:58:53.463004 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0330 09:58:53.463018 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14145 (* 1 = 1.14145 loss) | |
| I0330 09:58:53.463033 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.322802 (* 1 = 0.322802 loss) | |
| I0330 09:58:53.463045 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 09:58:53.463057 10583 solver.cpp:245] Train net output #16: total_confidence = 0.338147 | |
| I0330 09:58:53.463069 10583 sgd_solver.cpp:106] Iteration 122000, lr = 0.01 | |
| I0330 10:01:03.050185 10583 solver.cpp:229] Iteration 122500, loss = 3.03413 | |
| I0330 10:01:03.050344 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.307692 | |
| I0330 10:01:03.050365 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 10:01:03.050379 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.538462 | |
| I0330 10:01:03.050395 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5763 (* 0.3 = 0.772889 loss) | |
| I0330 10:01:03.050415 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.827956 (* 0.3 = 0.248387 loss) | |
| I0330 10:01:03.050428 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.365385 | |
| I0330 10:01:03.050441 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 10:01:03.050452 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.807692 | |
| I0330 10:01:03.050467 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.00388 (* 0.3 = 0.601165 loss) | |
| I0330 10:01:03.050480 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.628207 (* 0.3 = 0.188462 loss) | |
| I0330 10:01:03.050494 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.711538 | |
| I0330 10:01:03.050506 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 10:01:03.050518 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826923 | |
| I0330 10:01:03.050532 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28257 (* 1 = 1.28257 loss) | |
| I0330 10:01:03.050546 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.418851 (* 1 = 0.418851 loss) | |
| I0330 10:01:03.050559 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:01:03.050571 10583 solver.cpp:245] Train net output #16: total_confidence = 0.0894527 | |
| I0330 10:01:03.050585 10583 sgd_solver.cpp:106] Iteration 122500, lr = 0.01 | |
| I0330 10:03:12.530941 10583 solver.cpp:229] Iteration 123000, loss = 3.05381 | |
| I0330 10:03:12.531072 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431373 | |
| I0330 10:03:12.531093 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 10:03:12.531116 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 10:03:12.531131 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86354 (* 0.3 = 0.559061 loss) | |
| I0330 10:03:12.531147 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.573548 (* 0.3 = 0.172064 loss) | |
| I0330 10:03:12.531170 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54902 | |
| I0330 10:03:12.531184 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 10:03:12.531196 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72549 | |
| I0330 10:03:12.531210 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.62427 (* 0.3 = 0.487281 loss) | |
| I0330 10:03:12.531224 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.512505 (* 0.3 = 0.153751 loss) | |
| I0330 10:03:12.531236 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.745098 | |
| I0330 10:03:12.531250 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 10:03:12.531260 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.941176 | |
| I0330 10:03:12.531275 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.877268 (* 1 = 0.877268 loss) | |
| I0330 10:03:12.531288 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27756 (* 1 = 0.27756 loss) | |
| I0330 10:03:12.531301 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:03:12.531313 10583 solver.cpp:245] Train net output #16: total_confidence = 0.213561 | |
| I0330 10:03:12.531325 10583 sgd_solver.cpp:106] Iteration 123000, lr = 0.01 | |
| I0330 10:05:21.980865 10583 solver.cpp:229] Iteration 123500, loss = 2.97236 | |
| I0330 10:05:21.981019 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.36 | |
| I0330 10:05:21.981040 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 10:05:21.981055 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 10:05:21.981070 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41266 (* 0.3 = 0.723799 loss) | |
| I0330 10:05:21.981086 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.733964 (* 0.3 = 0.220189 loss) | |
| I0330 10:05:21.981102 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.42 | |
| I0330 10:05:21.981132 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 10:05:21.981145 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0330 10:05:21.981163 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79149 (* 0.3 = 0.537446 loss) | |
| I0330 10:05:21.981178 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.548099 (* 0.3 = 0.16443 loss) | |
| I0330 10:05:21.981190 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.68 | |
| I0330 10:05:21.981202 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 10:05:21.981215 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0330 10:05:21.981230 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.11867 (* 1 = 1.11867 loss) | |
| I0330 10:05:21.981243 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.338569 (* 1 = 0.338569 loss) | |
| I0330 10:05:21.981256 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:05:21.981268 10583 solver.cpp:245] Train net output #16: total_confidence = 0.244115 | |
| I0330 10:05:21.981281 10583 sgd_solver.cpp:106] Iteration 123500, lr = 0.01 | |
| I0330 10:07:31.327409 10583 solver.cpp:229] Iteration 124000, loss = 2.94495 | |
| I0330 10:07:31.327548 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.475 | |
| I0330 10:07:31.327569 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 10:07:31.327581 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0330 10:07:31.327600 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.71574 (* 0.3 = 0.514721 loss) | |
| I0330 10:07:31.327615 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.499026 (* 0.3 = 0.149708 loss) | |
| I0330 10:07:31.327627 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.65 | |
| I0330 10:07:31.327641 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 10:07:31.327652 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 10:07:31.327666 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.43979 (* 0.3 = 0.431936 loss) | |
| I0330 10:07:31.327680 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.459584 (* 0.3 = 0.137875 loss) | |
| I0330 10:07:31.327693 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.825 | |
| I0330 10:07:31.327705 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 10:07:31.327718 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 10:07:31.327733 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.760517 (* 1 = 0.760517 loss) | |
| I0330 10:07:31.327746 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.252711 (* 1 = 0.252711 loss) | |
| I0330 10:07:31.327759 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:07:31.327770 10583 solver.cpp:245] Train net output #16: total_confidence = 0.327284 | |
| I0330 10:07:31.327783 10583 sgd_solver.cpp:106] Iteration 124000, lr = 0.01 | |
| I0330 10:09:40.560583 10583 solver.cpp:229] Iteration 124500, loss = 2.92804 | |
| I0330 10:09:40.560734 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.283019 | |
| I0330 10:09:40.560755 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 10:09:40.560767 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54717 | |
| I0330 10:09:40.560784 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.27509 (* 0.3 = 0.682528 loss) | |
| I0330 10:09:40.560808 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.698624 (* 0.3 = 0.209587 loss) | |
| I0330 10:09:40.560820 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45283 | |
| I0330 10:09:40.560833 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 10:09:40.560845 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.735849 | |
| I0330 10:09:40.560859 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.78785 (* 0.3 = 0.536354 loss) | |
| I0330 10:09:40.560873 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.54869 (* 0.3 = 0.164607 loss) | |
| I0330 10:09:40.560886 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.716981 | |
| I0330 10:09:40.560897 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 10:09:40.560909 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.811321 | |
| I0330 10:09:40.560923 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03273 (* 1 = 1.03273 loss) | |
| I0330 10:09:40.560937 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.331804 (* 1 = 0.331804 loss) | |
| I0330 10:09:40.560950 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:09:40.560962 10583 solver.cpp:245] Train net output #16: total_confidence = 0.309757 | |
| I0330 10:09:40.560974 10583 sgd_solver.cpp:106] Iteration 124500, lr = 0.01 | |
| I0330 10:11:49.875787 10583 solver.cpp:338] Iteration 125000, Testing net (#0) | |
| I0330 10:12:19.692747 10583 solver.cpp:393] Test loss: 279.48 | |
| I0330 10:12:19.692801 10583 solver.cpp:406] Test net output #0: loss1/accuracy = 0 | |
| I0330 10:12:19.692818 10583 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.758955 | |
| I0330 10:12:19.692831 10583 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0 | |
| I0330 10:12:19.692848 10583 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 10:12:19.692864 10583 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 10:12:19.692878 10583 solver.cpp:406] Test net output #5: loss2/accuracy = 0 | |
| I0330 10:12:19.692889 10583 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.758955 | |
| I0330 10:12:19.692900 10583 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0 | |
| I0330 10:12:19.692914 10583 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 10:12:19.692930 10583 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 87.3361 (* 0.3 = 26.2008 loss) | |
| I0330 10:12:19.692942 10583 solver.cpp:406] Test net output #10: loss3/accuracy = 0 | |
| I0330 10:12:19.692955 10583 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.758955 | |
| I0330 10:12:19.692966 10583 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0 | |
| I0330 10:12:19.692981 10583 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 10:12:19.692996 10583 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 87.3361 (* 1 = 87.3361 loss) | |
| I0330 10:12:19.693007 10583 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0330 10:12:19.693019 10583 solver.cpp:406] Test net output #16: total_confidence = nan | |
| I0330 10:12:19.844372 10583 solver.cpp:229] Iteration 125000, loss = 2.9701 | |
| I0330 10:12:19.844415 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.348837 | |
| I0330 10:12:19.844432 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 10:12:19.844445 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.697674 | |
| I0330 10:12:19.844465 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.951 (* 0.3 = 0.5853 loss) | |
| I0330 10:12:19.844491 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.525049 (* 0.3 = 0.157515 loss) | |
| I0330 10:12:19.844511 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.465116 | |
| I0330 10:12:19.844523 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 10:12:19.844535 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.790698 | |
| I0330 10:12:19.844549 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53 (* 0.3 = 0.458999 loss) | |
| I0330 10:12:19.844563 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.428183 (* 0.3 = 0.128455 loss) | |
| I0330 10:12:19.844580 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.744186 | |
| I0330 10:12:19.844594 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 10:12:19.844605 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.837209 | |
| I0330 10:12:19.844619 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.953324 (* 1 = 0.953324 loss) | |
| I0330 10:12:19.844633 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.250124 (* 1 = 0.250124 loss) | |
| I0330 10:12:19.844645 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 10:12:19.844657 10583 solver.cpp:245] Train net output #16: total_confidence = 0.248437 | |
| I0330 10:12:19.844671 10583 sgd_solver.cpp:106] Iteration 125000, lr = 0.01 | |
| I0330 10:14:29.140928 10583 solver.cpp:229] Iteration 125500, loss = 3.0164 | |
| I0330 10:14:29.141089 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0330 10:14:29.141110 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 10:14:29.141124 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.714286 | |
| I0330 10:14:29.141139 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.71728 (* 0.3 = 0.515183 loss) | |
| I0330 10:14:29.141155 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.633768 (* 0.3 = 0.19013 loss) | |
| I0330 10:14:29.141170 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.642857 | |
| I0330 10:14:29.141182 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 10:14:29.141194 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0330 10:14:29.141208 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41489 (* 0.3 = 0.424467 loss) | |
| I0330 10:14:29.141222 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.492296 (* 0.3 = 0.147689 loss) | |
| I0330 10:14:29.141234 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.809524 | |
| I0330 10:14:29.141247 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 10:14:29.141258 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97619 | |
| I0330 10:14:29.141273 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.624602 (* 1 = 0.624602 loss) | |
| I0330 10:14:29.141286 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.200953 (* 1 = 0.200953 loss) | |
| I0330 10:14:29.141299 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:14:29.141310 10583 solver.cpp:245] Train net output #16: total_confidence = 0.318026 | |
| I0330 10:14:29.141324 10583 sgd_solver.cpp:106] Iteration 125500, lr = 0.01 | |
| I0330 10:16:38.576058 10583 solver.cpp:229] Iteration 126000, loss = 2.86701 | |
| I0330 10:16:38.576213 10583 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 10:16:38.576242 10583 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 10:16:38.576256 10583 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 10:16:38.576272 10583 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.38588 (* 0.3 = 0.715763 loss) | |
| I0330 10:16:38.576287 10583 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.700824 (* 0.3 = 0.210247 loss) | |
| I0330 10:16:38.576299 10583 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377778 | |
| I0330 10:16:38.576313 10583 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 10:16:38.576323 10583 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.688889 | |
| I0330 10:16:38.576339 10583 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75117 (* 0.3 = 0.525352 loss) | |
| I0330 10:16:38.576354 10583 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.503318 (* 0.3 = 0.150995 loss) | |
| I0330 10:16:38.576365 10583 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 10:16:38.576377 10583 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 10:16:38.576390 10583 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.911111 | |
| I0330 10:16:38.576405 10583 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.09973 (* 1 = 1.09973 loss) | |
| I0330 10:16:38.576418 10583 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.34801 (* 1 = 0.34801 loss) | |
| I0330 10:16:38.576438 10583 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:16:38.576450 10583 solver.cpp:245] Train net output #16: total_confidence = 0.217483 | |
| I0330 10:16:38.576462 10583 sgd_solver.cpp:106] Iteration 126000, lr = 0.01 | |
| I0330 10:17:08.792840 10583 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_126117.caffemodel | |
| I0330 10:17:09.125098 10583 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_126117.solverstate | |
| I0330 10:17:09.293496 10583 solver.cpp:302] Optimization stopped early. | |
| I0330 10:17:09.293555 10583 caffe.cpp:222] Optimization Done. | |
| I0330 10:32:44.919842 13762 solver.cpp:280] Solving mixed_lstm | |
| I0330 10:32:44.919855 13762 solver.cpp:281] Learning Rate Policy: fixed | |
| I0330 10:32:44.940197 13762 solver.cpp:338] Iteration 0, Testing net (#0) | |
| I0330 10:33:14.970613 13762 solver.cpp:393] Test loss: 2.42662 | |
| I0330 10:33:14.971014 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.436481 | |
| I0330 10:33:14.971035 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.857684 | |
| I0330 10:33:14.971047 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.72661 | |
| I0330 10:33:14.971063 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.8831 (* 0.3 = 0.56493 loss) | |
| I0330 10:33:14.971078 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.485121 (* 0.3 = 0.145536 loss) | |
| I0330 10:33:14.971091 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.61077 | |
| I0330 10:33:14.971103 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.897821 | |
| I0330 10:33:14.971114 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.842172 | |
| I0330 10:33:14.971128 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.34895 (* 0.3 = 0.404685 loss) | |
| I0330 10:33:14.971143 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.356009 (* 0.3 = 0.106803 loss) | |
| I0330 10:33:14.971154 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.7628 | |
| I0330 10:33:14.971169 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.932683 | |
| I0330 10:33:14.971181 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.883967 | |
| I0330 10:33:14.971195 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.941209 (* 1 = 0.941209 loss) | |
| I0330 10:33:14.971210 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.263462 (* 1 = 0.263462 loss) | |
| I0330 10:33:14.971220 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.356 | |
| I0330 10:33:14.971232 13762 solver.cpp:406] Test net output #16: total_confidence = 0.294884 | |
| I0330 10:33:15.278595 13762 solver.cpp:229] Iteration 0, loss = 3.9675 | |
| I0330 10:33:15.278650 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0330 10:33:15.278667 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 10:33:15.278681 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.466667 | |
| I0330 10:33:15.278699 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.56424 (* 0.3 = 0.769272 loss) | |
| I0330 10:33:15.278715 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.709118 (* 0.3 = 0.212735 loss) | |
| I0330 10:33:15.278728 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 10:33:15.278741 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 10:33:15.278753 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 10:33:15.278775 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.90732 (* 0.3 = 0.572196 loss) | |
| I0330 10:33:15.278790 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.554373 (* 0.3 = 0.166312 loss) | |
| I0330 10:33:15.278802 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.511111 | |
| I0330 10:33:15.278815 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 10:33:15.278826 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.733333 | |
| I0330 10:33:15.278846 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.73724 (* 1 = 1.73724 loss) | |
| I0330 10:33:15.278859 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.509744 (* 1 = 0.509744 loss) | |
| I0330 10:33:15.278872 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:33:15.278884 13762 solver.cpp:245] Train net output #16: total_confidence = 0.0669346 | |
| I0330 10:33:15.278903 13762 sgd_solver.cpp:106] Iteration 0, lr = 0.01 | |
| I0330 10:35:23.937925 13762 solver.cpp:229] Iteration 500, loss = 2.95395 | |
| I0330 10:35:23.938068 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.36 | |
| I0330 10:35:23.938088 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 10:35:23.938102 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0330 10:35:23.938125 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.0097 (* 0.3 = 0.602911 loss) | |
| I0330 10:35:23.938141 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.621908 (* 0.3 = 0.186572 loss) | |
| I0330 10:35:23.938154 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44 | |
| I0330 10:35:23.938169 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 10:35:23.938182 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.74 | |
| I0330 10:35:23.938196 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.59861 (* 0.3 = 0.479583 loss) | |
| I0330 10:35:23.938211 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.478211 (* 0.3 = 0.143463 loss) | |
| I0330 10:35:23.938225 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.68 | |
| I0330 10:35:23.938236 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 10:35:23.938248 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.82 | |
| I0330 10:35:23.938262 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3527 (* 1 = 1.3527 loss) | |
| I0330 10:35:23.938277 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.391665 (* 1 = 0.391665 loss) | |
| I0330 10:35:23.938288 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 10:35:23.938302 13762 solver.cpp:245] Train net output #16: total_confidence = 0.377938 | |
| I0330 10:35:23.938321 13762 sgd_solver.cpp:106] Iteration 500, lr = 0.01 | |
| I0330 10:37:32.562679 13762 solver.cpp:229] Iteration 1000, loss = 2.93511 | |
| I0330 10:37:32.562805 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 10:37:32.562824 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 10:37:32.562844 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509804 | |
| I0330 10:37:32.562860 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.1964 (* 0.3 = 0.658919 loss) | |
| I0330 10:37:32.562875 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.657657 (* 0.3 = 0.197297 loss) | |
| I0330 10:37:32.562887 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.470588 | |
| I0330 10:37:32.562907 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 10:37:32.562921 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.803922 | |
| I0330 10:37:32.562934 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.5393 (* 0.3 = 0.46179 loss) | |
| I0330 10:37:32.562948 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.469256 (* 0.3 = 0.140777 loss) | |
| I0330 10:37:32.562961 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.72549 | |
| I0330 10:37:32.562988 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 10:37:32.563004 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.843137 | |
| I0330 10:37:32.563017 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.1011 (* 1 = 1.1011 loss) | |
| I0330 10:37:32.563031 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.348225 (* 1 = 0.348225 loss) | |
| I0330 10:37:32.563045 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:37:32.563056 13762 solver.cpp:245] Train net output #16: total_confidence = 0.289157 | |
| I0330 10:37:32.563068 13762 sgd_solver.cpp:106] Iteration 1000, lr = 0.01 | |
| I0330 10:39:41.266428 13762 solver.cpp:229] Iteration 1500, loss = 2.91519 | |
| I0330 10:39:41.266600 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0330 10:39:41.266633 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 10:39:41.266646 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55 | |
| I0330 10:39:41.266664 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25479 (* 0.3 = 0.676437 loss) | |
| I0330 10:39:41.266680 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.614997 (* 0.3 = 0.184499 loss) | |
| I0330 10:39:41.266691 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.35 | |
| I0330 10:39:41.266705 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 10:39:41.266717 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.625 | |
| I0330 10:39:41.266731 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.91465 (* 0.3 = 0.574396 loss) | |
| I0330 10:39:41.266746 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524363 (* 0.3 = 0.157309 loss) | |
| I0330 10:39:41.266757 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6 | |
| I0330 10:39:41.266769 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 10:39:41.266782 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.775 | |
| I0330 10:39:41.266796 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.58724 (* 1 = 1.58724 loss) | |
| I0330 10:39:41.266810 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.47902 (* 1 = 0.47902 loss) | |
| I0330 10:39:41.266827 13762 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 10:39:41.266839 13762 solver.cpp:245] Train net output #16: total_confidence = 0.228045 | |
| I0330 10:39:41.266850 13762 sgd_solver.cpp:106] Iteration 1500, lr = 0.01 | |
| I0330 10:41:49.853971 13762 solver.cpp:229] Iteration 2000, loss = 2.94361 | |
| I0330 10:41:49.854094 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0330 10:41:49.854115 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 10:41:49.854127 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0330 10:41:49.854149 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5854 (* 0.3 = 0.775621 loss) | |
| I0330 10:41:49.854166 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.781194 (* 0.3 = 0.234358 loss) | |
| I0330 10:41:49.854181 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4375 | |
| I0330 10:41:49.854199 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 10:41:49.854212 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6875 | |
| I0330 10:41:49.854225 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12723 (* 0.3 = 0.638168 loss) | |
| I0330 10:41:49.854239 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.664462 (* 0.3 = 0.199339 loss) | |
| I0330 10:41:49.854252 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.583333 | |
| I0330 10:41:49.854264 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 10:41:49.854276 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.708333 | |
| I0330 10:41:49.854291 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.7779 (* 1 = 1.7779 loss) | |
| I0330 10:41:49.854305 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.543329 (* 1 = 0.543329 loss) | |
| I0330 10:41:49.854317 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:41:49.854329 13762 solver.cpp:245] Train net output #16: total_confidence = 0.106113 | |
| I0330 10:41:49.854342 13762 sgd_solver.cpp:106] Iteration 2000, lr = 0.01 | |
| I0330 10:43:58.487763 13762 solver.cpp:229] Iteration 2500, loss = 2.9616 | |
| I0330 10:43:58.487905 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 10:43:58.487925 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 10:43:58.487937 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0330 10:43:58.487960 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.7502 (* 0.3 = 0.525061 loss) | |
| I0330 10:43:58.487975 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.521847 (* 0.3 = 0.156554 loss) | |
| I0330 10:43:58.487988 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0330 10:43:58.488001 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 10:43:58.488014 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 10:43:58.488029 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.57513 (* 0.3 = 0.47254 loss) | |
| I0330 10:43:58.488044 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.451091 (* 0.3 = 0.135327 loss) | |
| I0330 10:43:58.488056 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.775 | |
| I0330 10:43:58.488068 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 10:43:58.488080 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0330 10:43:58.488095 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.752328 (* 1 = 0.752328 loss) | |
| I0330 10:43:58.488111 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.218702 (* 1 = 0.218702 loss) | |
| I0330 10:43:58.488122 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 10:43:58.488134 13762 solver.cpp:245] Train net output #16: total_confidence = 0.230088 | |
| I0330 10:43:58.488147 13762 sgd_solver.cpp:106] Iteration 2500, lr = 0.01 | |
| I0330 10:46:06.977277 13762 solver.cpp:229] Iteration 3000, loss = 2.91295 | |
| I0330 10:46:06.977383 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 10:46:06.977403 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 10:46:06.977416 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 10:46:06.977432 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96901 (* 0.3 = 0.590702 loss) | |
| I0330 10:46:06.977447 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559247 (* 0.3 = 0.167774 loss) | |
| I0330 10:46:06.977460 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.533333 | |
| I0330 10:46:06.977473 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 10:46:06.977485 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.822222 | |
| I0330 10:46:06.977499 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41369 (* 0.3 = 0.424106 loss) | |
| I0330 10:46:06.977514 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.403716 (* 0.3 = 0.121115 loss) | |
| I0330 10:46:06.977527 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0330 10:46:06.977540 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 10:46:06.977551 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0330 10:46:06.977568 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.651035 (* 1 = 0.651035 loss) | |
| I0330 10:46:06.977583 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.169315 (* 1 = 0.169315 loss) | |
| I0330 10:46:06.977596 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:46:06.977608 13762 solver.cpp:245] Train net output #16: total_confidence = 0.378488 | |
| I0330 10:46:06.977620 13762 sgd_solver.cpp:106] Iteration 3000, lr = 0.01 | |
| I0330 10:48:15.997232 13762 solver.cpp:229] Iteration 3500, loss = 2.96821 | |
| I0330 10:48:15.997385 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.575 | |
| I0330 10:48:15.997407 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.875 | |
| I0330 10:48:15.997421 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.775 | |
| I0330 10:48:15.997437 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.45492 (* 0.3 = 0.436475 loss) | |
| I0330 10:48:15.997452 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.438254 (* 0.3 = 0.131476 loss) | |
| I0330 10:48:15.997465 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.725 | |
| I0330 10:48:15.997478 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0330 10:48:15.997490 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.875 | |
| I0330 10:48:15.997505 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.04696 (* 0.3 = 0.314087 loss) | |
| I0330 10:48:15.997519 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.328799 (* 0.3 = 0.0986397 loss) | |
| I0330 10:48:15.997532 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.875 | |
| I0330 10:48:15.997545 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 10:48:15.997557 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 10:48:15.997571 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.660786 (* 1 = 0.660786 loss) | |
| I0330 10:48:15.997586 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.172294 (* 1 = 0.172294 loss) | |
| I0330 10:48:15.997598 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 10:48:15.997611 13762 solver.cpp:245] Train net output #16: total_confidence = 0.457124 | |
| I0330 10:48:15.997622 13762 sgd_solver.cpp:106] Iteration 3500, lr = 0.01 | |
| I0330 10:50:24.488608 13762 solver.cpp:229] Iteration 4000, loss = 2.90772 | |
| I0330 10:50:24.488721 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.358974 | |
| I0330 10:50:24.488740 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 10:50:24.488754 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.641026 | |
| I0330 10:50:24.488770 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14433 (* 0.3 = 0.643299 loss) | |
| I0330 10:50:24.488785 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.61285 (* 0.3 = 0.183855 loss) | |
| I0330 10:50:24.488797 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.487179 | |
| I0330 10:50:24.488809 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 10:50:24.488822 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.769231 | |
| I0330 10:50:24.488837 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65883 (* 0.3 = 0.49765 loss) | |
| I0330 10:50:24.488852 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.478937 (* 0.3 = 0.143681 loss) | |
| I0330 10:50:24.488863 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.871795 | |
| I0330 10:50:24.488875 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 10:50:24.488888 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0330 10:50:24.488903 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.771457 (* 1 = 0.771457 loss) | |
| I0330 10:50:24.488916 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.207043 (* 1 = 0.207043 loss) | |
| I0330 10:50:24.488929 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 10:50:24.488941 13762 solver.cpp:245] Train net output #16: total_confidence = 0.208056 | |
| I0330 10:50:24.488953 13762 sgd_solver.cpp:106] Iteration 4000, lr = 0.01 | |
| I0330 10:52:33.133946 13762 solver.cpp:229] Iteration 4500, loss = 2.8923 | |
| I0330 10:52:33.134090 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395349 | |
| I0330 10:52:33.134110 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 10:52:33.134132 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.767442 | |
| I0330 10:52:33.134148 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.75898 (* 0.3 = 0.527693 loss) | |
| I0330 10:52:33.134166 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.500696 (* 0.3 = 0.150209 loss) | |
| I0330 10:52:33.134179 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.581395 | |
| I0330 10:52:33.134192 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 10:52:33.134205 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.883721 | |
| I0330 10:52:33.134227 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.2704 (* 0.3 = 0.381121 loss) | |
| I0330 10:52:33.134263 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.360241 (* 0.3 = 0.108072 loss) | |
| I0330 10:52:33.134289 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.930233 | |
| I0330 10:52:33.134315 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0330 10:52:33.134327 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 10:52:33.134341 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.330809 (* 1 = 0.330809 loss) | |
| I0330 10:52:33.134356 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0976903 (* 1 = 0.0976903 loss) | |
| I0330 10:52:33.134369 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 10:52:33.134382 13762 solver.cpp:245] Train net output #16: total_confidence = 0.301994 | |
| I0330 10:52:33.134393 13762 sgd_solver.cpp:106] Iteration 4500, lr = 0.01 | |
| I0330 10:54:41.662323 13762 solver.cpp:338] Iteration 5000, Testing net (#0) | |
| I0330 10:55:11.575417 13762 solver.cpp:393] Test loss: 2.48033 | |
| I0330 10:55:11.575469 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.477805 | |
| I0330 10:55:11.575496 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.846457 | |
| I0330 10:55:11.575520 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.752162 | |
| I0330 10:55:11.575547 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.75037 (* 0.3 = 0.525111 loss) | |
| I0330 10:55:11.575574 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.513095 (* 0.3 = 0.153928 loss) | |
| I0330 10:55:11.575598 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.622059 | |
| I0330 10:55:11.575623 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.897775 | |
| I0330 10:55:11.575646 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.844068 | |
| I0330 10:55:11.575671 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.34068 (* 0.3 = 0.402205 loss) | |
| I0330 10:55:11.575698 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.358835 (* 0.3 = 0.107651 loss) | |
| I0330 10:55:11.575721 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.746315 | |
| I0330 10:55:11.575742 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.93241 | |
| I0330 10:55:11.575762 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.874356 | |
| I0330 10:55:11.575788 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.02164 (* 1 = 1.02164 loss) | |
| I0330 10:55:11.575814 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.269791 (* 1 = 0.269791 loss) | |
| I0330 10:55:11.575834 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.382 | |
| I0330 10:55:11.575856 13762 solver.cpp:406] Test net output #16: total_confidence = 0.33548 | |
| I0330 10:55:11.728579 13762 solver.cpp:229] Iteration 5000, loss = 2.92707 | |
| I0330 10:55:11.728770 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.36 | |
| I0330 10:55:11.728798 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 10:55:11.728821 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 10:55:11.728848 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12923 (* 0.3 = 0.63877 loss) | |
| I0330 10:55:11.728876 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.626342 (* 0.3 = 0.187903 loss) | |
| I0330 10:55:11.728899 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.34 | |
| I0330 10:55:11.728922 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 10:55:11.728945 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.66 | |
| I0330 10:55:11.728971 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95148 (* 0.3 = 0.585445 loss) | |
| I0330 10:55:11.728997 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.576533 (* 0.3 = 0.17296 loss) | |
| I0330 10:55:11.729020 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.66 | |
| I0330 10:55:11.729045 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 10:55:11.729069 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0330 10:55:11.729096 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.18722 (* 1 = 1.18722 loss) | |
| I0330 10:55:11.729123 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.347903 (* 1 = 0.347903 loss) | |
| I0330 10:55:11.729146 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:55:11.729173 13762 solver.cpp:245] Train net output #16: total_confidence = 0.224876 | |
| I0330 10:55:11.729197 13762 sgd_solver.cpp:106] Iteration 5000, lr = 0.01 | |
| I0330 10:57:20.400054 13762 solver.cpp:229] Iteration 5500, loss = 2.91833 | |
| I0330 10:57:20.400162 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.24 | |
| I0330 10:57:20.400182 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 10:57:20.400195 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.52 | |
| I0330 10:57:20.400212 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48154 (* 0.3 = 0.744463 loss) | |
| I0330 10:57:20.400226 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.733707 (* 0.3 = 0.220112 loss) | |
| I0330 10:57:20.400239 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.42 | |
| I0330 10:57:20.400252 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 10:57:20.400264 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0330 10:57:20.400279 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.98116 (* 0.3 = 0.594348 loss) | |
| I0330 10:57:20.400293 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.592646 (* 0.3 = 0.177794 loss) | |
| I0330 10:57:20.400306 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0330 10:57:20.400318 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 10:57:20.400331 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.82 | |
| I0330 10:57:20.400344 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.05647 (* 1 = 1.05647 loss) | |
| I0330 10:57:20.400359 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.33196 (* 1 = 0.33196 loss) | |
| I0330 10:57:20.400372 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 10:57:20.400383 13762 solver.cpp:245] Train net output #16: total_confidence = 0.119639 | |
| I0330 10:57:20.400395 13762 sgd_solver.cpp:106] Iteration 5500, lr = 0.01 | |
| I0330 10:59:29.093415 13762 solver.cpp:229] Iteration 6000, loss = 2.87537 | |
| I0330 10:59:29.093552 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0330 10:59:29.093582 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 10:59:29.093618 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.428571 | |
| I0330 10:59:29.093646 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.50484 (* 0.3 = 0.751451 loss) | |
| I0330 10:59:29.093662 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.784398 (* 0.3 = 0.23532 loss) | |
| I0330 10:59:29.093675 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.367347 | |
| I0330 10:59:29.093688 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 10:59:29.093703 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.632653 | |
| I0330 10:59:29.093718 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.15363 (* 0.3 = 0.646088 loss) | |
| I0330 10:59:29.093732 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.675971 (* 0.3 = 0.202791 loss) | |
| I0330 10:59:29.093744 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0330 10:59:29.093765 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 10:59:29.093777 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795918 | |
| I0330 10:59:29.093791 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28904 (* 1 = 1.28904 loss) | |
| I0330 10:59:29.093806 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.413978 (* 1 = 0.413978 loss) | |
| I0330 10:59:29.093817 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 10:59:29.093829 13762 solver.cpp:245] Train net output #16: total_confidence = 0.199179 | |
| I0330 10:59:29.093842 13762 sgd_solver.cpp:106] Iteration 6000, lr = 0.01 | |
| I0330 11:01:37.701958 13762 solver.cpp:229] Iteration 6500, loss = 2.90704 | |
| I0330 11:01:37.702080 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.416667 | |
| I0330 11:01:37.702100 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 11:01:37.702112 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6875 | |
| I0330 11:01:37.702128 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02157 (* 0.3 = 0.606472 loss) | |
| I0330 11:01:37.702143 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.629904 (* 0.3 = 0.188971 loss) | |
| I0330 11:01:37.702157 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 11:01:37.702172 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 11:01:37.702184 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0330 11:01:37.702199 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.56484 (* 0.3 = 0.469453 loss) | |
| I0330 11:01:37.702214 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.485881 (* 0.3 = 0.145764 loss) | |
| I0330 11:01:37.702226 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.729167 | |
| I0330 11:01:37.702239 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 11:01:37.702257 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 11:01:37.702275 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.05774 (* 1 = 1.05774 loss) | |
| I0330 11:01:37.702288 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.316883 (* 1 = 0.316883 loss) | |
| I0330 11:01:37.702301 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:01:37.702312 13762 solver.cpp:245] Train net output #16: total_confidence = 0.348862 | |
| I0330 11:01:37.702329 13762 sgd_solver.cpp:106] Iteration 6500, lr = 0.01 | |
| I0330 11:03:46.176281 13762 solver.cpp:229] Iteration 7000, loss = 2.87881 | |
| I0330 11:03:46.176405 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.230769 | |
| I0330 11:03:46.176424 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 11:03:46.176437 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.538462 | |
| I0330 11:03:46.176462 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42261 (* 0.3 = 0.726784 loss) | |
| I0330 11:03:46.176477 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.749409 (* 0.3 = 0.224823 loss) | |
| I0330 11:03:46.176491 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.487179 | |
| I0330 11:03:46.176502 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 11:03:46.176514 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.692308 | |
| I0330 11:03:46.176528 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.99231 (* 0.3 = 0.597694 loss) | |
| I0330 11:03:46.176542 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.624057 (* 0.3 = 0.187217 loss) | |
| I0330 11:03:46.176555 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 11:03:46.176568 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 11:03:46.176579 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.769231 | |
| I0330 11:03:46.176594 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.38971 (* 1 = 1.38971 loss) | |
| I0330 11:03:46.176607 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.392445 (* 1 = 0.392445 loss) | |
| I0330 11:03:46.176620 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:03:46.176632 13762 solver.cpp:245] Train net output #16: total_confidence = 0.36541 | |
| I0330 11:03:46.176645 13762 sgd_solver.cpp:106] Iteration 7000, lr = 0.01 | |
| I0330 11:05:54.739506 13762 solver.cpp:229] Iteration 7500, loss = 2.87713 | |
| I0330 11:05:54.739624 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.342105 | |
| I0330 11:05:54.739644 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 11:05:54.739657 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.605263 | |
| I0330 11:05:54.739673 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.16269 (* 0.3 = 0.648807 loss) | |
| I0330 11:05:54.739688 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.628445 (* 0.3 = 0.188533 loss) | |
| I0330 11:05:54.739701 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.631579 | |
| I0330 11:05:54.739714 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 11:05:54.739727 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.815789 | |
| I0330 11:05:54.739742 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.38817 (* 0.3 = 0.416451 loss) | |
| I0330 11:05:54.739756 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.437115 (* 0.3 = 0.131134 loss) | |
| I0330 11:05:54.739768 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.789474 | |
| I0330 11:05:54.739781 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 11:05:54.739799 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.921053 | |
| I0330 11:05:54.739814 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.658469 (* 1 = 0.658469 loss) | |
| I0330 11:05:54.739827 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.2227 (* 1 = 0.2227 loss) | |
| I0330 11:05:54.739840 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:05:54.739861 13762 solver.cpp:245] Train net output #16: total_confidence = 0.301178 | |
| I0330 11:05:54.739873 13762 sgd_solver.cpp:106] Iteration 7500, lr = 0.01 | |
| I0330 11:08:03.396921 13762 solver.cpp:229] Iteration 8000, loss = 2.82707 | |
| I0330 11:08:03.397090 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 11:08:03.397111 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:08:03.397132 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.690476 | |
| I0330 11:08:03.397148 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.98276 (* 0.3 = 0.594827 loss) | |
| I0330 11:08:03.397166 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.598089 (* 0.3 = 0.179427 loss) | |
| I0330 11:08:03.397179 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.47619 | |
| I0330 11:08:03.397192 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 11:08:03.397205 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.809524 | |
| I0330 11:08:03.397219 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.42402 (* 0.3 = 0.427205 loss) | |
| I0330 11:08:03.397233 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.464868 (* 0.3 = 0.13946 loss) | |
| I0330 11:08:03.397246 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.690476 | |
| I0330 11:08:03.397258 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 11:08:03.397270 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0330 11:08:03.397285 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.920011 (* 1 = 0.920011 loss) | |
| I0330 11:08:03.397300 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.25396 (* 1 = 0.25396 loss) | |
| I0330 11:08:03.397311 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:08:03.397325 13762 solver.cpp:245] Train net output #16: total_confidence = 0.248094 | |
| I0330 11:08:03.397341 13762 sgd_solver.cpp:106] Iteration 8000, lr = 0.01 | |
| I0330 11:10:12.149786 13762 solver.cpp:229] Iteration 8500, loss = 2.89731 | |
| I0330 11:10:12.149912 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27451 | |
| I0330 11:10:12.149932 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 11:10:12.149945 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.607843 | |
| I0330 11:10:12.149961 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.62976 (* 0.3 = 0.788928 loss) | |
| I0330 11:10:12.149976 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.847046 (* 0.3 = 0.254114 loss) | |
| I0330 11:10:12.149989 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.411765 | |
| I0330 11:10:12.150002 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 11:10:12.150014 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 11:10:12.150028 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.01732 (* 0.3 = 0.605195 loss) | |
| I0330 11:10:12.150043 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.656554 (* 0.3 = 0.196966 loss) | |
| I0330 11:10:12.150055 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.745098 | |
| I0330 11:10:12.150068 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 11:10:12.150079 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.784314 | |
| I0330 11:10:12.150094 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.25472 (* 1 = 1.25472 loss) | |
| I0330 11:10:12.150109 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.431271 (* 1 = 0.431271 loss) | |
| I0330 11:10:12.150121 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:10:12.150133 13762 solver.cpp:245] Train net output #16: total_confidence = 0.264847 | |
| I0330 11:10:12.150146 13762 sgd_solver.cpp:106] Iteration 8500, lr = 0.01 | |
| I0330 11:12:20.943611 13762 solver.cpp:229] Iteration 9000, loss = 2.94884 | |
| I0330 11:12:20.943738 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.418605 | |
| I0330 11:12:20.943758 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 11:12:20.943771 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.674419 | |
| I0330 11:12:20.943795 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.88254 (* 0.3 = 0.564761 loss) | |
| I0330 11:12:20.943810 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.542492 (* 0.3 = 0.162748 loss) | |
| I0330 11:12:20.943824 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.651163 | |
| I0330 11:12:20.943835 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 11:12:20.943848 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.860465 | |
| I0330 11:12:20.943861 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32965 (* 0.3 = 0.398894 loss) | |
| I0330 11:12:20.943876 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.430848 (* 0.3 = 0.129255 loss) | |
| I0330 11:12:20.943888 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.813953 | |
| I0330 11:12:20.943907 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 11:12:20.943920 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.976744 | |
| I0330 11:12:20.943934 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.550532 (* 1 = 0.550532 loss) | |
| I0330 11:12:20.943948 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.188393 (* 1 = 0.188393 loss) | |
| I0330 11:12:20.943960 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:12:20.943972 13762 solver.cpp:245] Train net output #16: total_confidence = 0.466632 | |
| I0330 11:12:20.943984 13762 sgd_solver.cpp:106] Iteration 9000, lr = 0.01 | |
| I0330 11:14:29.632549 13762 solver.cpp:229] Iteration 9500, loss = 2.90313 | |
| I0330 11:14:29.632645 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.294118 | |
| I0330 11:14:29.632663 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 11:14:29.632676 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.568627 | |
| I0330 11:14:29.632693 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.52478 (* 0.3 = 0.757433 loss) | |
| I0330 11:14:29.632707 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.789889 (* 0.3 = 0.236967 loss) | |
| I0330 11:14:29.632719 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.294118 | |
| I0330 11:14:29.632735 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0330 11:14:29.632748 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.568627 | |
| I0330 11:14:29.632762 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.35139 (* 0.3 = 0.705416 loss) | |
| I0330 11:14:29.632777 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.707483 (* 0.3 = 0.212245 loss) | |
| I0330 11:14:29.632789 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0330 11:14:29.632802 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 11:14:29.632813 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.843137 | |
| I0330 11:14:29.632828 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26276 (* 1 = 1.26276 loss) | |
| I0330 11:14:29.632843 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.43522 (* 1 = 0.43522 loss) | |
| I0330 11:14:29.632854 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 11:14:29.632866 13762 solver.cpp:245] Train net output #16: total_confidence = 0.0802933 | |
| I0330 11:14:29.632879 13762 sgd_solver.cpp:106] Iteration 9500, lr = 0.01 | |
| I0330 11:16:38.117166 13762 solver.cpp:338] Iteration 10000, Testing net (#0) | |
| I0330 11:17:08.121039 13762 solver.cpp:393] Test loss: 2.59024 | |
| I0330 11:17:08.121160 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.468171 | |
| I0330 11:17:08.121179 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.862958 | |
| I0330 11:17:08.121192 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.747068 | |
| I0330 11:17:08.121208 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.82894 (* 0.3 = 0.548682 loss) | |
| I0330 11:17:08.121222 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.479253 (* 0.3 = 0.143776 loss) | |
| I0330 11:17:08.121235 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.587243 | |
| I0330 11:17:08.121248 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.89382 | |
| I0330 11:17:08.121259 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.829284 | |
| I0330 11:17:08.121273 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.4761 (* 0.3 = 0.442831 loss) | |
| I0330 11:17:08.121286 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.383423 (* 0.3 = 0.115027 loss) | |
| I0330 11:17:08.121299 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.745473 | |
| I0330 11:17:08.121310 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.933364 | |
| I0330 11:17:08.121322 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.875427 | |
| I0330 11:17:08.121335 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.06288 (* 1 = 1.06288 loss) | |
| I0330 11:17:08.121350 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.277048 (* 1 = 0.277048 loss) | |
| I0330 11:17:08.121361 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.371 | |
| I0330 11:17:08.121372 13762 solver.cpp:406] Test net output #16: total_confidence = 0.312618 | |
| I0330 11:17:08.273020 13762 solver.cpp:229] Iteration 10000, loss = 2.91491 | |
| I0330 11:17:08.273059 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3 | |
| I0330 11:17:08.273077 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:17:08.273089 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55 | |
| I0330 11:17:08.273104 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.31043 (* 0.3 = 0.69313 loss) | |
| I0330 11:17:08.273119 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.653307 (* 0.3 = 0.195992 loss) | |
| I0330 11:17:08.273131 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.475 | |
| I0330 11:17:08.273144 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 11:17:08.273156 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.675 | |
| I0330 11:17:08.273169 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.71197 (* 0.3 = 0.513591 loss) | |
| I0330 11:17:08.273185 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.470822 (* 0.3 = 0.141247 loss) | |
| I0330 11:17:08.273196 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.725 | |
| I0330 11:17:08.273208 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 11:17:08.273221 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 11:17:08.273234 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.07187 (* 1 = 1.07187 loss) | |
| I0330 11:17:08.273248 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.277804 (* 1 = 0.277804 loss) | |
| I0330 11:17:08.273260 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 11:17:08.273272 13762 solver.cpp:245] Train net output #16: total_confidence = 0.194258 | |
| I0330 11:17:08.273285 13762 sgd_solver.cpp:106] Iteration 10000, lr = 0.01 | |
| I0330 11:19:17.004987 13762 solver.cpp:229] Iteration 10500, loss = 2.85151 | |
| I0330 11:19:17.005131 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.306122 | |
| I0330 11:19:17.005151 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:19:17.005173 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0330 11:19:17.005192 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.28786 (* 0.3 = 0.686358 loss) | |
| I0330 11:19:17.005206 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.648577 (* 0.3 = 0.194573 loss) | |
| I0330 11:19:17.005219 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.469388 | |
| I0330 11:19:17.005233 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 11:19:17.005245 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0330 11:19:17.005259 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65884 (* 0.3 = 0.497653 loss) | |
| I0330 11:19:17.005273 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.491803 (* 0.3 = 0.147541 loss) | |
| I0330 11:19:17.005286 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.77551 | |
| I0330 11:19:17.005298 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 11:19:17.005311 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0330 11:19:17.005332 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.910061 (* 1 = 0.910061 loss) | |
| I0330 11:19:17.005347 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.265437 (* 1 = 0.265437 loss) | |
| I0330 11:19:17.005359 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:19:17.005372 13762 solver.cpp:245] Train net output #16: total_confidence = 0.414926 | |
| I0330 11:19:17.005383 13762 sgd_solver.cpp:106] Iteration 10500, lr = 0.01 | |
| I0330 11:21:25.698953 13762 solver.cpp:229] Iteration 11000, loss = 2.83528 | |
| I0330 11:21:25.699098 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 11:21:25.699120 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 11:21:25.699132 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.44898 | |
| I0330 11:21:25.699149 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32662 (* 0.3 = 0.697987 loss) | |
| I0330 11:21:25.699167 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.702856 (* 0.3 = 0.210857 loss) | |
| I0330 11:21:25.699180 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489796 | |
| I0330 11:21:25.699193 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 11:21:25.699205 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673469 | |
| I0330 11:21:25.699219 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.78882 (* 0.3 = 0.536645 loss) | |
| I0330 11:21:25.699234 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.558254 (* 0.3 = 0.167476 loss) | |
| I0330 11:21:25.699246 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.530612 | |
| I0330 11:21:25.699259 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 11:21:25.699271 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0330 11:21:25.699285 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.473 (* 1 = 1.473 loss) | |
| I0330 11:21:25.699301 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.445199 (* 1 = 0.445199 loss) | |
| I0330 11:21:25.699312 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 11:21:25.699324 13762 solver.cpp:245] Train net output #16: total_confidence = 0.23561 | |
| I0330 11:21:25.699337 13762 sgd_solver.cpp:106] Iteration 11000, lr = 0.01 | |
| I0330 11:23:34.753463 13762 solver.cpp:229] Iteration 11500, loss = 2.91294 | |
| I0330 11:23:34.753623 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.352941 | |
| I0330 11:23:34.753662 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:23:34.753686 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54902 | |
| I0330 11:23:34.753715 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.0148 (* 0.3 = 0.90444 loss) | |
| I0330 11:23:34.753743 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.910264 (* 0.3 = 0.273079 loss) | |
| I0330 11:23:34.753767 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45098 | |
| I0330 11:23:34.753790 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 11:23:34.753815 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.686275 | |
| I0330 11:23:34.753842 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.7261 (* 0.3 = 0.817829 loss) | |
| I0330 11:23:34.753870 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.826523 (* 0.3 = 0.247957 loss) | |
| I0330 11:23:34.753895 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.745098 | |
| I0330 11:23:34.753937 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 11:23:34.753958 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.803922 | |
| I0330 11:23:34.753991 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.85223 (* 1 = 1.85223 loss) | |
| I0330 11:23:34.754017 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.552092 (* 1 = 0.552092 loss) | |
| I0330 11:23:34.754041 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:23:34.754062 13762 solver.cpp:245] Train net output #16: total_confidence = 0.361682 | |
| I0330 11:23:34.754084 13762 sgd_solver.cpp:106] Iteration 11500, lr = 0.01 | |
| I0330 11:25:43.585342 13762 solver.cpp:229] Iteration 12000, loss = 2.81827 | |
| I0330 11:25:43.585455 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.311111 | |
| I0330 11:25:43.585475 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 11:25:43.585489 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.466667 | |
| I0330 11:25:43.585505 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.6064 (* 0.3 = 0.781921 loss) | |
| I0330 11:25:43.585520 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.77136 (* 0.3 = 0.231408 loss) | |
| I0330 11:25:43.585533 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 11:25:43.585546 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 11:25:43.585558 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0330 11:25:43.585573 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.08014 (* 0.3 = 0.624043 loss) | |
| I0330 11:25:43.585587 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.645826 (* 0.3 = 0.193748 loss) | |
| I0330 11:25:43.585600 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0330 11:25:43.585613 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 11:25:43.585625 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.711111 | |
| I0330 11:25:43.585639 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.58097 (* 1 = 1.58097 loss) | |
| I0330 11:25:43.585654 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.510785 (* 1 = 0.510785 loss) | |
| I0330 11:25:43.585665 13762 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0330 11:25:43.585677 13762 solver.cpp:245] Train net output #16: total_confidence = 0.187881 | |
| I0330 11:25:43.585690 13762 sgd_solver.cpp:106] Iteration 12000, lr = 0.01 | |
| I0330 11:27:52.574936 13762 solver.cpp:229] Iteration 12500, loss = 2.83511 | |
| I0330 11:27:52.575104 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.386364 | |
| I0330 11:27:52.575134 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 11:27:52.575170 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.545455 | |
| I0330 11:27:52.575199 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.47097 (* 0.3 = 0.741292 loss) | |
| I0330 11:27:52.575227 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.69369 (* 0.3 = 0.208107 loss) | |
| I0330 11:27:52.575251 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.454545 | |
| I0330 11:27:52.575276 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 11:27:52.575300 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.681818 | |
| I0330 11:27:52.575326 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06996 (* 0.3 = 0.620987 loss) | |
| I0330 11:27:52.575361 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.576692 (* 0.3 = 0.173008 loss) | |
| I0330 11:27:52.575387 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.522727 | |
| I0330 11:27:52.575408 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 11:27:52.575430 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795455 | |
| I0330 11:27:52.575456 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60453 (* 1 = 1.60453 loss) | |
| I0330 11:27:52.575482 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.425388 (* 1 = 0.425388 loss) | |
| I0330 11:27:52.575505 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 11:27:52.575525 13762 solver.cpp:245] Train net output #16: total_confidence = 0.174991 | |
| I0330 11:27:52.575547 13762 sgd_solver.cpp:106] Iteration 12500, lr = 0.01 | |
| I0330 11:30:01.363342 13762 solver.cpp:229] Iteration 13000, loss = 2.90868 | |
| I0330 11:30:01.363445 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 11:30:01.363466 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 11:30:01.363478 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.458333 | |
| I0330 11:30:01.363494 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70935 (* 0.3 = 0.812806 loss) | |
| I0330 11:30:01.363509 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.769831 (* 0.3 = 0.230949 loss) | |
| I0330 11:30:01.363523 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0330 11:30:01.363534 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0330 11:30:01.363548 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.541667 | |
| I0330 11:30:01.363561 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.40968 (* 0.3 = 0.722904 loss) | |
| I0330 11:30:01.363575 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.752566 (* 0.3 = 0.22577 loss) | |
| I0330 11:30:01.363589 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.541667 | |
| I0330 11:30:01.363600 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 11:30:01.363612 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.625 | |
| I0330 11:30:01.363627 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.74591 (* 1 = 1.74591 loss) | |
| I0330 11:30:01.363641 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.560636 (* 1 = 0.560636 loss) | |
| I0330 11:30:01.363653 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 11:30:01.363665 13762 solver.cpp:245] Train net output #16: total_confidence = 0.170615 | |
| I0330 11:30:01.363677 13762 sgd_solver.cpp:106] Iteration 13000, lr = 0.01 | |
| I0330 11:32:10.240480 13762 solver.cpp:229] Iteration 13500, loss = 2.83737 | |
| I0330 11:32:10.240725 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.288462 | |
| I0330 11:32:10.240747 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 11:32:10.240761 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.653846 | |
| I0330 11:32:10.240779 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15971 (* 0.3 = 0.647913 loss) | |
| I0330 11:32:10.240794 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.647316 (* 0.3 = 0.194195 loss) | |
| I0330 11:32:10.240816 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.384615 | |
| I0330 11:32:10.240829 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 11:32:10.240844 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.711538 | |
| I0330 11:32:10.240857 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.74719 (* 0.3 = 0.524158 loss) | |
| I0330 11:32:10.240880 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.545343 (* 0.3 = 0.163603 loss) | |
| I0330 11:32:10.240892 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 11:32:10.240906 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 11:32:10.240917 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0330 11:32:10.240932 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.758314 (* 1 = 0.758314 loss) | |
| I0330 11:32:10.240947 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.235384 (* 1 = 0.235384 loss) | |
| I0330 11:32:10.240959 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:32:10.240972 13762 solver.cpp:245] Train net output #16: total_confidence = 0.3711 | |
| I0330 11:32:10.240984 13762 sgd_solver.cpp:106] Iteration 13500, lr = 0.01 | |
| I0330 11:34:19.178539 13762 solver.cpp:229] Iteration 14000, loss = 2.86182 | |
| I0330 11:34:19.178660 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0330 11:34:19.178681 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 11:34:19.178694 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0330 11:34:19.178710 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93459 (* 0.3 = 0.580376 loss) | |
| I0330 11:34:19.178725 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.617874 (* 0.3 = 0.185362 loss) | |
| I0330 11:34:19.178738 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0330 11:34:19.178750 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 11:34:19.178763 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0330 11:34:19.178777 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66666 (* 0.3 = 0.499997 loss) | |
| I0330 11:34:19.178791 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.510691 (* 0.3 = 0.153207 loss) | |
| I0330 11:34:19.178804 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.636364 | |
| I0330 11:34:19.178817 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 11:34:19.178828 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.772727 | |
| I0330 11:34:19.178843 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.13607 (* 1 = 1.13607 loss) | |
| I0330 11:34:19.178858 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.364312 (* 1 = 0.364312 loss) | |
| I0330 11:34:19.178869 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:34:19.178882 13762 solver.cpp:245] Train net output #16: total_confidence = 0.466338 | |
| I0330 11:34:19.178894 13762 sgd_solver.cpp:106] Iteration 14000, lr = 0.01 | |
| I0330 11:36:28.201479 13762 solver.cpp:229] Iteration 14500, loss = 2.76228 | |
| I0330 11:36:28.201670 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.434783 | |
| I0330 11:36:28.201710 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 11:36:28.201735 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0330 11:36:28.201764 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41274 (* 0.3 = 0.723821 loss) | |
| I0330 11:36:28.201792 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.767538 (* 0.3 = 0.230261 loss) | |
| I0330 11:36:28.201813 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.586957 | |
| I0330 11:36:28.201836 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 11:36:28.201860 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0330 11:36:28.201887 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.58959 (* 0.3 = 0.476878 loss) | |
| I0330 11:36:28.201915 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532919 (* 0.3 = 0.159876 loss) | |
| I0330 11:36:28.201937 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.652174 | |
| I0330 11:36:28.201978 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 11:36:28.202002 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804348 | |
| I0330 11:36:28.202035 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.15813 (* 1 = 1.15813 loss) | |
| I0330 11:36:28.202060 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.366462 (* 1 = 0.366462 loss) | |
| I0330 11:36:28.202090 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 11:36:28.202111 13762 solver.cpp:245] Train net output #16: total_confidence = 0.303602 | |
| I0330 11:36:28.202133 13762 sgd_solver.cpp:106] Iteration 14500, lr = 0.01 | |
| I0330 11:38:37.288952 13762 solver.cpp:338] Iteration 15000, Testing net (#0) | |
| I0330 11:39:07.414701 13762 solver.cpp:393] Test loss: 2.49485 | |
| I0330 11:39:07.414832 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.49582 | |
| I0330 11:39:07.414860 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.854594 | |
| I0330 11:39:07.414891 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.771927 | |
| I0330 11:39:07.414916 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.72067 (* 0.3 = 0.516202 loss) | |
| I0330 11:39:07.414942 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.501013 (* 0.3 = 0.150304 loss) | |
| I0330 11:39:07.414963 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.627491 | |
| I0330 11:39:07.415002 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.894048 | |
| I0330 11:39:07.415029 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.848989 | |
| I0330 11:39:07.415055 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.31613 (* 0.3 = 0.394838 loss) | |
| I0330 11:39:07.415079 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.378354 (* 0.3 = 0.113506 loss) | |
| I0330 11:39:07.415099 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.730141 | |
| I0330 11:39:07.415122 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.925046 | |
| I0330 11:39:07.415144 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.874934 | |
| I0330 11:39:07.415175 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.02831 (* 1 = 1.02831 loss) | |
| I0330 11:39:07.415202 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.291691 (* 1 = 0.291691 loss) | |
| I0330 11:39:07.415225 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.344 | |
| I0330 11:39:07.415252 13762 solver.cpp:406] Test net output #16: total_confidence = 0.277267 | |
| I0330 11:39:07.566823 13762 solver.cpp:229] Iteration 15000, loss = 2.96408 | |
| I0330 11:39:07.566879 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.346939 | |
| I0330 11:39:07.566905 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:39:07.566951 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0330 11:39:07.566984 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.30823 (* 0.3 = 0.692468 loss) | |
| I0330 11:39:07.567013 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.720547 (* 0.3 = 0.216164 loss) | |
| I0330 11:39:07.567037 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510204 | |
| I0330 11:39:07.567062 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 11:39:07.567085 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.653061 | |
| I0330 11:39:07.567111 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9814 (* 0.3 = 0.594419 loss) | |
| I0330 11:39:07.567140 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.61458 (* 0.3 = 0.184374 loss) | |
| I0330 11:39:07.567165 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.693878 | |
| I0330 11:39:07.567188 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 11:39:07.567211 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.877551 | |
| I0330 11:39:07.567237 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24006 (* 1 = 1.24006 loss) | |
| I0330 11:39:07.567263 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.454275 (* 1 = 0.454275 loss) | |
| I0330 11:39:07.567286 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 11:39:07.567307 13762 solver.cpp:245] Train net output #16: total_confidence = 0.301619 | |
| I0330 11:39:07.567329 13762 sgd_solver.cpp:106] Iteration 15000, lr = 0.01 | |
| I0330 11:41:16.996129 13762 solver.cpp:229] Iteration 15500, loss = 2.84135 | |
| I0330 11:41:16.996340 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.511111 | |
| I0330 11:41:16.996363 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0330 11:41:16.996376 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.755556 | |
| I0330 11:41:16.996393 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86337 (* 0.3 = 0.559012 loss) | |
| I0330 11:41:16.996408 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.543938 (* 0.3 = 0.163181 loss) | |
| I0330 11:41:16.996422 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.422222 | |
| I0330 11:41:16.996434 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 11:41:16.996448 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0330 11:41:16.996461 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67218 (* 0.3 = 0.501655 loss) | |
| I0330 11:41:16.996476 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.508856 (* 0.3 = 0.152657 loss) | |
| I0330 11:41:16.996490 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0330 11:41:16.996501 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 11:41:16.996513 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0330 11:41:16.996528 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.700152 (* 1 = 0.700152 loss) | |
| I0330 11:41:16.996542 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.197003 (* 1 = 0.197003 loss) | |
| I0330 11:41:16.996556 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 11:41:16.996567 13762 solver.cpp:245] Train net output #16: total_confidence = 0.493137 | |
| I0330 11:41:16.996580 13762 sgd_solver.cpp:106] Iteration 15500, lr = 0.01 | |
| I0330 11:43:27.241900 13762 solver.cpp:229] Iteration 16000, loss = 2.77562 | |
| I0330 11:43:27.242123 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.348837 | |
| I0330 11:43:27.242151 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 11:43:27.242167 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.651163 | |
| I0330 11:43:27.242185 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09471 (* 0.3 = 0.628413 loss) | |
| I0330 11:43:27.242209 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.641642 (* 0.3 = 0.192493 loss) | |
| I0330 11:43:27.242223 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.581395 | |
| I0330 11:43:27.242235 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 11:43:27.242247 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.813953 | |
| I0330 11:43:27.242267 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66247 (* 0.3 = 0.498742 loss) | |
| I0330 11:43:27.242281 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.523337 (* 0.3 = 0.157001 loss) | |
| I0330 11:43:27.242295 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.813953 | |
| I0330 11:43:27.242307 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 11:43:27.242319 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0330 11:43:27.242334 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.33891 (* 1 = 1.33891 loss) | |
| I0330 11:43:27.242348 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.371185 (* 1 = 0.371185 loss) | |
| I0330 11:43:27.242362 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:43:27.242373 13762 solver.cpp:245] Train net output #16: total_confidence = 0.232614 | |
| I0330 11:43:27.242388 13762 sgd_solver.cpp:106] Iteration 16000, lr = 0.01 | |
| I0330 11:45:37.277298 13762 solver.cpp:229] Iteration 16500, loss = 2.84836 | |
| I0330 11:45:37.277451 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0330 11:45:37.277472 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 11:45:37.277484 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.466667 | |
| I0330 11:45:37.277500 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25598 (* 0.3 = 0.676795 loss) | |
| I0330 11:45:37.277515 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.600253 (* 0.3 = 0.180076 loss) | |
| I0330 11:45:37.277529 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.266667 | |
| I0330 11:45:37.277542 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 11:45:37.277555 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.633333 | |
| I0330 11:45:37.277570 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96303 (* 0.3 = 0.588909 loss) | |
| I0330 11:45:37.277585 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.555879 (* 0.3 = 0.166764 loss) | |
| I0330 11:45:37.277598 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0330 11:45:37.277611 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 11:45:37.277623 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 11:45:37.277638 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.11336 (* 1 = 1.11336 loss) | |
| I0330 11:45:37.277652 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.271489 (* 1 = 0.271489 loss) | |
| I0330 11:45:37.277673 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 11:45:37.277685 13762 solver.cpp:245] Train net output #16: total_confidence = 0.122483 | |
| I0330 11:45:37.277698 13762 sgd_solver.cpp:106] Iteration 16500, lr = 0.01 | |
| I0330 11:47:47.407516 13762 solver.cpp:229] Iteration 17000, loss = 2.93393 | |
| I0330 11:47:47.407727 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.245283 | |
| I0330 11:47:47.407758 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 11:47:47.407780 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.584906 | |
| I0330 11:47:47.407807 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70535 (* 0.3 = 0.811606 loss) | |
| I0330 11:47:47.407835 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.834938 (* 0.3 = 0.250481 loss) | |
| I0330 11:47:47.407858 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.415094 | |
| I0330 11:47:47.407881 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 11:47:47.407902 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.716981 | |
| I0330 11:47:47.407928 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03385 (* 0.3 = 0.610154 loss) | |
| I0330 11:47:47.407954 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.626355 (* 0.3 = 0.187906 loss) | |
| I0330 11:47:47.407975 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.660377 | |
| I0330 11:47:47.407999 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 11:47:47.408023 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.849057 | |
| I0330 11:47:47.408051 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.10398 (* 1 = 1.10398 loss) | |
| I0330 11:47:47.408077 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.368522 (* 1 = 0.368522 loss) | |
| I0330 11:47:47.408099 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 11:47:47.408130 13762 solver.cpp:245] Train net output #16: total_confidence = 0.134158 | |
| I0330 11:47:47.408155 13762 sgd_solver.cpp:106] Iteration 17000, lr = 0.01 | |
| I0330 11:49:57.335924 13762 solver.cpp:229] Iteration 17500, loss = 2.79844 | |
| I0330 11:49:57.336094 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0330 11:49:57.336125 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 11:49:57.336155 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0330 11:49:57.336187 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07577 (* 0.3 = 0.622732 loss) | |
| I0330 11:49:57.336215 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.570988 (* 0.3 = 0.171296 loss) | |
| I0330 11:49:57.336244 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0330 11:49:57.336266 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 11:49:57.336288 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.697674 | |
| I0330 11:49:57.336314 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79168 (* 0.3 = 0.537505 loss) | |
| I0330 11:49:57.336341 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.520749 (* 0.3 = 0.156225 loss) | |
| I0330 11:49:57.336364 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.744186 | |
| I0330 11:49:57.336385 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 11:49:57.336408 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.860465 | |
| I0330 11:49:57.336437 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.869522 (* 1 = 0.869522 loss) | |
| I0330 11:49:57.336464 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.217681 (* 1 = 0.217681 loss) | |
| I0330 11:49:57.336488 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 11:49:57.336508 13762 solver.cpp:245] Train net output #16: total_confidence = 0.346285 | |
| I0330 11:49:57.336549 13762 sgd_solver.cpp:106] Iteration 17500, lr = 0.01 | |
| I0330 11:52:07.374969 13762 solver.cpp:229] Iteration 18000, loss = 2.7677 | |
| I0330 11:52:07.375216 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.358491 | |
| I0330 11:52:07.375247 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 11:52:07.375277 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.603774 | |
| I0330 11:52:07.375304 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17796 (* 0.3 = 0.653388 loss) | |
| I0330 11:52:07.375339 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.676347 (* 0.3 = 0.202904 loss) | |
| I0330 11:52:07.375360 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.490566 | |
| I0330 11:52:07.375383 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 11:52:07.375406 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.754717 | |
| I0330 11:52:07.375432 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.73888 (* 0.3 = 0.521666 loss) | |
| I0330 11:52:07.375458 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.533597 (* 0.3 = 0.160079 loss) | |
| I0330 11:52:07.375478 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.773585 | |
| I0330 11:52:07.375501 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 11:52:07.375524 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.849057 | |
| I0330 11:52:07.375553 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.951202 (* 1 = 0.951202 loss) | |
| I0330 11:52:07.375581 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.303779 (* 1 = 0.303779 loss) | |
| I0330 11:52:07.375603 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:52:07.375624 13762 solver.cpp:245] Train net output #16: total_confidence = 0.18175 | |
| I0330 11:52:07.375653 13762 sgd_solver.cpp:106] Iteration 18000, lr = 0.01 | |
| I0330 11:54:16.881804 13762 solver.cpp:229] Iteration 18500, loss = 2.81778 | |
| I0330 11:54:16.881958 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.327273 | |
| I0330 11:54:16.881980 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 11:54:16.881994 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.563636 | |
| I0330 11:54:16.882011 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.57198 (* 0.3 = 0.771593 loss) | |
| I0330 11:54:16.882026 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.817248 (* 0.3 = 0.245174 loss) | |
| I0330 11:54:16.882040 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.418182 | |
| I0330 11:54:16.882052 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 11:54:16.882066 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.563636 | |
| I0330 11:54:16.882079 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.33672 (* 0.3 = 0.701016 loss) | |
| I0330 11:54:16.882094 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.742654 (* 0.3 = 0.222796 loss) | |
| I0330 11:54:16.882107 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.527273 | |
| I0330 11:54:16.882120 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 11:54:16.882133 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.636364 | |
| I0330 11:54:16.882148 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.84535 (* 1 = 1.84535 loss) | |
| I0330 11:54:16.882164 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.59517 (* 1 = 0.59517 loss) | |
| I0330 11:54:16.882177 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 11:54:16.882189 13762 solver.cpp:245] Train net output #16: total_confidence = 0.195179 | |
| I0330 11:54:16.882202 13762 sgd_solver.cpp:106] Iteration 18500, lr = 0.01 | |
| I0330 11:56:26.069244 13762 solver.cpp:229] Iteration 19000, loss = 2.83997 | |
| I0330 11:56:26.069429 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.469388 | |
| I0330 11:56:26.069452 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 11:56:26.069471 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.673469 | |
| I0330 11:56:26.069489 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.85959 (* 0.3 = 0.557877 loss) | |
| I0330 11:56:26.069504 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.532762 (* 0.3 = 0.159829 loss) | |
| I0330 11:56:26.069516 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.653061 | |
| I0330 11:56:26.069528 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0330 11:56:26.069541 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.795918 | |
| I0330 11:56:26.069555 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.34431 (* 0.3 = 0.403294 loss) | |
| I0330 11:56:26.069571 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.39567 (* 0.3 = 0.118701 loss) | |
| I0330 11:56:26.069583 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.795918 | |
| I0330 11:56:26.069596 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 11:56:26.069607 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.918367 | |
| I0330 11:56:26.069622 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.673661 (* 1 = 0.673661 loss) | |
| I0330 11:56:26.069635 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.207512 (* 1 = 0.207512 loss) | |
| I0330 11:56:26.069648 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 11:56:26.069664 13762 solver.cpp:245] Train net output #16: total_confidence = 0.313088 | |
| I0330 11:56:26.069686 13762 sgd_solver.cpp:106] Iteration 19000, lr = 0.01 | |
| I0330 11:58:35.245337 13762 solver.cpp:229] Iteration 19500, loss = 2.77294 | |
| I0330 11:58:35.245486 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 11:58:35.245507 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 11:58:35.245519 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.642857 | |
| I0330 11:58:35.245537 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07919 (* 0.3 = 0.623758 loss) | |
| I0330 11:58:35.245550 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.655604 (* 0.3 = 0.196681 loss) | |
| I0330 11:58:35.245563 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.547619 | |
| I0330 11:58:35.245576 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 11:58:35.245589 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.809524 | |
| I0330 11:58:35.245602 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.50673 (* 0.3 = 0.452019 loss) | |
| I0330 11:58:35.245617 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.45021 (* 0.3 = 0.135063 loss) | |
| I0330 11:58:35.245630 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0330 11:58:35.245642 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 11:58:35.245654 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.904762 | |
| I0330 11:58:35.245668 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.895394 (* 1 = 0.895394 loss) | |
| I0330 11:58:35.245682 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.230626 (* 1 = 0.230626 loss) | |
| I0330 11:58:35.245694 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 11:58:35.245707 13762 solver.cpp:245] Train net output #16: total_confidence = 0.281781 | |
| I0330 11:58:35.245720 13762 sgd_solver.cpp:106] Iteration 19500, lr = 0.01 | |
| I0330 12:00:44.246582 13762 solver.cpp:338] Iteration 20000, Testing net (#0) | |
| I0330 12:01:14.403830 13762 solver.cpp:393] Test loss: 2.34307 | |
| I0330 12:01:14.403944 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.510317 | |
| I0330 12:01:14.403964 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.864049 | |
| I0330 12:01:14.403977 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.770174 | |
| I0330 12:01:14.403993 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.68078 (* 0.3 = 0.504233 loss) | |
| I0330 12:01:14.404008 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.46806 (* 0.3 = 0.140418 loss) | |
| I0330 12:01:14.404021 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.637123 | |
| I0330 12:01:14.404032 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.905003 | |
| I0330 12:01:14.404044 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.856929 | |
| I0330 12:01:14.404059 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.25854 (* 0.3 = 0.377562 loss) | |
| I0330 12:01:14.404074 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.331862 (* 0.3 = 0.0995585 loss) | |
| I0330 12:01:14.404086 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.761915 | |
| I0330 12:01:14.404098 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.938955 | |
| I0330 12:01:14.404110 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.884064 | |
| I0330 12:01:14.404124 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.969504 (* 1 = 0.969504 loss) | |
| I0330 12:01:14.404139 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.251794 (* 1 = 0.251794 loss) | |
| I0330 12:01:14.404150 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.423 | |
| I0330 12:01:14.404172 13762 solver.cpp:406] Test net output #16: total_confidence = 0.422184 | |
| I0330 12:01:14.557221 13762 solver.cpp:229] Iteration 20000, loss = 2.82634 | |
| I0330 12:01:14.557282 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.387755 | |
| I0330 12:01:14.557299 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 12:01:14.557312 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0330 12:01:14.557329 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.7124 (* 0.3 = 0.813719 loss) | |
| I0330 12:01:14.557344 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.854735 (* 0.3 = 0.25642 loss) | |
| I0330 12:01:14.557356 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55102 | |
| I0330 12:01:14.557369 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 12:01:14.557381 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.734694 | |
| I0330 12:01:14.557396 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06785 (* 0.3 = 0.620355 loss) | |
| I0330 12:01:14.557410 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.645143 (* 0.3 = 0.193543 loss) | |
| I0330 12:01:14.557423 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.632653 | |
| I0330 12:01:14.557437 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 12:01:14.557449 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.77551 | |
| I0330 12:01:14.557463 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.8492 (* 1 = 1.8492 loss) | |
| I0330 12:01:14.557477 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.555984 (* 1 = 0.555984 loss) | |
| I0330 12:01:14.557489 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:01:14.557502 13762 solver.cpp:245] Train net output #16: total_confidence = 0.235477 | |
| I0330 12:01:14.557514 13762 sgd_solver.cpp:106] Iteration 20000, lr = 0.01 | |
| I0330 12:03:23.870220 13762 solver.cpp:229] Iteration 20500, loss = 2.7444 | |
| I0330 12:03:23.870373 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.510638 | |
| I0330 12:03:23.870394 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0330 12:03:23.870407 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.765957 | |
| I0330 12:03:23.870424 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.66149 (* 0.3 = 0.498446 loss) | |
| I0330 12:03:23.870437 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.476154 (* 0.3 = 0.142846 loss) | |
| I0330 12:03:23.870450 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.446809 | |
| I0330 12:03:23.870465 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 12:03:23.870476 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.87234 | |
| I0330 12:03:23.870496 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41306 (* 0.3 = 0.423918 loss) | |
| I0330 12:03:23.870509 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.419143 (* 0.3 = 0.125743 loss) | |
| I0330 12:03:23.870522 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829787 | |
| I0330 12:03:23.870533 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 12:03:23.870545 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.957447 | |
| I0330 12:03:23.870559 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.648965 (* 1 = 0.648965 loss) | |
| I0330 12:03:23.870573 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.183082 (* 1 = 0.183082 loss) | |
| I0330 12:03:23.870585 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:03:23.870597 13762 solver.cpp:245] Train net output #16: total_confidence = 0.220974 | |
| I0330 12:03:23.870609 13762 sgd_solver.cpp:106] Iteration 20500, lr = 0.01 | |
| I0330 12:05:33.758486 13762 solver.cpp:229] Iteration 21000, loss = 2.89182 | |
| I0330 12:05:33.758621 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.422222 | |
| I0330 12:05:33.758642 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 12:05:33.758654 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.711111 | |
| I0330 12:05:33.758671 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.88857 (* 0.3 = 0.566572 loss) | |
| I0330 12:05:33.758685 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.51525 (* 0.3 = 0.154575 loss) | |
| I0330 12:05:33.758698 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0330 12:05:33.758710 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 12:05:33.758723 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0330 12:05:33.758736 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65673 (* 0.3 = 0.497019 loss) | |
| I0330 12:05:33.758750 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.460687 (* 0.3 = 0.138206 loss) | |
| I0330 12:05:33.758764 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.755556 | |
| I0330 12:05:33.758775 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 12:05:33.758787 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0330 12:05:33.758801 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.07263 (* 1 = 1.07263 loss) | |
| I0330 12:05:33.758816 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.278552 (* 1 = 0.278552 loss) | |
| I0330 12:05:33.758834 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:05:33.758846 13762 solver.cpp:245] Train net output #16: total_confidence = 0.326774 | |
| I0330 12:05:33.758858 13762 sgd_solver.cpp:106] Iteration 21000, lr = 0.01 | |
| I0330 12:07:42.995929 13762 solver.cpp:229] Iteration 21500, loss = 2.78979 | |
| I0330 12:07:42.996140 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.46875 | |
| I0330 12:07:42.996163 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0330 12:07:42.996178 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.59375 | |
| I0330 12:07:42.996203 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.88246 (* 0.3 = 0.564739 loss) | |
| I0330 12:07:42.996219 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.521107 (* 0.3 = 0.156332 loss) | |
| I0330 12:07:42.996232 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.59375 | |
| I0330 12:07:42.996245 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 12:07:42.996258 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.78125 | |
| I0330 12:07:42.996273 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.21436 (* 0.3 = 0.364309 loss) | |
| I0330 12:07:42.996286 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.381027 (* 0.3 = 0.114308 loss) | |
| I0330 12:07:42.996300 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.65625 | |
| I0330 12:07:42.996312 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 12:07:42.996325 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.84375 | |
| I0330 12:07:42.996340 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.06362 (* 1 = 1.06362 loss) | |
| I0330 12:07:42.996356 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.310871 (* 1 = 0.310871 loss) | |
| I0330 12:07:42.996367 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 12:07:42.996381 13762 solver.cpp:245] Train net output #16: total_confidence = 0.190339 | |
| I0330 12:07:42.996393 13762 sgd_solver.cpp:106] Iteration 21500, lr = 0.01 | |
| I0330 12:09:52.288084 13762 solver.cpp:229] Iteration 22000, loss = 2.85292 | |
| I0330 12:09:52.288228 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488889 | |
| I0330 12:09:52.288249 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 12:09:52.288262 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 12:09:52.288280 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.94846 (* 0.3 = 0.584537 loss) | |
| I0330 12:09:52.288295 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.571479 (* 0.3 = 0.171444 loss) | |
| I0330 12:09:52.288306 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0330 12:09:52.288319 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 12:09:52.288332 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0330 12:09:52.288347 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88337 (* 0.3 = 0.56501 loss) | |
| I0330 12:09:52.288360 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.534764 (* 0.3 = 0.160429 loss) | |
| I0330 12:09:52.288374 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.688889 | |
| I0330 12:09:52.288386 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 12:09:52.288399 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.822222 | |
| I0330 12:09:52.288414 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.04357 (* 1 = 1.04357 loss) | |
| I0330 12:09:52.288429 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.301313 (* 1 = 0.301313 loss) | |
| I0330 12:09:52.288440 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:09:52.288453 13762 solver.cpp:245] Train net output #16: total_confidence = 0.314489 | |
| I0330 12:09:52.288466 13762 sgd_solver.cpp:106] Iteration 22000, lr = 0.01 | |
| I0330 12:12:01.418839 13762 solver.cpp:229] Iteration 22500, loss = 2.78587 | |
| I0330 12:12:01.419062 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.446809 | |
| I0330 12:12:01.419092 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 12:12:01.419106 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.723404 | |
| I0330 12:12:01.419123 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86234 (* 0.3 = 0.558701 loss) | |
| I0330 12:12:01.419138 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.554667 (* 0.3 = 0.1664 loss) | |
| I0330 12:12:01.419152 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.574468 | |
| I0330 12:12:01.419174 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 12:12:01.419188 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.787234 | |
| I0330 12:12:01.419201 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.50761 (* 0.3 = 0.452283 loss) | |
| I0330 12:12:01.419215 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.431763 (* 0.3 = 0.129529 loss) | |
| I0330 12:12:01.419229 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.702128 | |
| I0330 12:12:01.419240 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 12:12:01.419253 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.829787 | |
| I0330 12:12:01.419267 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.17007 (* 1 = 1.17007 loss) | |
| I0330 12:12:01.419281 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.353562 (* 1 = 0.353562 loss) | |
| I0330 12:12:01.419294 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:12:01.419306 13762 solver.cpp:245] Train net output #16: total_confidence = 0.308607 | |
| I0330 12:12:01.419319 13762 sgd_solver.cpp:106] Iteration 22500, lr = 0.01 | |
| I0330 12:14:10.674485 13762 solver.cpp:229] Iteration 23000, loss = 2.80221 | |
| I0330 12:14:10.674626 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.346939 | |
| I0330 12:14:10.674648 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 12:14:10.674660 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0330 12:14:10.674676 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.44383 (* 0.3 = 0.733148 loss) | |
| I0330 12:14:10.674691 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.820165 (* 0.3 = 0.24605 loss) | |
| I0330 12:14:10.674705 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489796 | |
| I0330 12:14:10.674717 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 12:14:10.674729 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0330 12:14:10.674744 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.78555 (* 0.3 = 0.535666 loss) | |
| I0330 12:14:10.674758 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.600011 (* 0.3 = 0.180003 loss) | |
| I0330 12:14:10.674772 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.734694 | |
| I0330 12:14:10.674790 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 12:14:10.674803 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.897959 | |
| I0330 12:14:10.674816 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.919039 (* 1 = 0.919039 loss) | |
| I0330 12:14:10.674830 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.355764 (* 1 = 0.355764 loss) | |
| I0330 12:14:10.674847 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 12:14:10.674860 13762 solver.cpp:245] Train net output #16: total_confidence = 0.1066 | |
| I0330 12:14:10.674872 13762 sgd_solver.cpp:106] Iteration 23000, lr = 0.01 | |
| I0330 12:16:20.263416 13762 solver.cpp:229] Iteration 23500, loss = 2.78545 | |
| I0330 12:16:20.263661 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.320755 | |
| I0330 12:16:20.263689 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 12:16:20.263718 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.528302 | |
| I0330 12:16:20.263736 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32774 (* 0.3 = 0.698321 loss) | |
| I0330 12:16:20.263751 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.735835 (* 0.3 = 0.220751 loss) | |
| I0330 12:16:20.263764 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.471698 | |
| I0330 12:16:20.263777 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 12:16:20.263789 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.773585 | |
| I0330 12:16:20.263804 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79255 (* 0.3 = 0.537765 loss) | |
| I0330 12:16:20.263818 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.570271 (* 0.3 = 0.171081 loss) | |
| I0330 12:16:20.263831 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.660377 | |
| I0330 12:16:20.263844 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 12:16:20.263855 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.886792 | |
| I0330 12:16:20.263870 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03523 (* 1 = 1.03523 loss) | |
| I0330 12:16:20.263885 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.328994 (* 1 = 0.328994 loss) | |
| I0330 12:16:20.263896 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:16:20.263909 13762 solver.cpp:245] Train net output #16: total_confidence = 0.134987 | |
| I0330 12:16:20.263922 13762 sgd_solver.cpp:106] Iteration 23500, lr = 0.01 | |
| I0330 12:18:29.876345 13762 solver.cpp:229] Iteration 24000, loss = 2.78313 | |
| I0330 12:18:29.876514 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 12:18:29.876536 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 12:18:29.876549 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510204 | |
| I0330 12:18:29.876567 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36391 (* 0.3 = 0.709172 loss) | |
| I0330 12:18:29.876582 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.706923 (* 0.3 = 0.212077 loss) | |
| I0330 12:18:29.876595 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.469388 | |
| I0330 12:18:29.876608 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 12:18:29.876621 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.734694 | |
| I0330 12:18:29.876636 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.74116 (* 0.3 = 0.522347 loss) | |
| I0330 12:18:29.876651 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.5073 (* 0.3 = 0.15219 loss) | |
| I0330 12:18:29.876662 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.632653 | |
| I0330 12:18:29.876675 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 12:18:29.876688 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.836735 | |
| I0330 12:18:29.876703 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34699 (* 1 = 1.34699 loss) | |
| I0330 12:18:29.876716 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.38573 (* 1 = 0.38573 loss) | |
| I0330 12:18:29.876729 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 12:18:29.876741 13762 solver.cpp:245] Train net output #16: total_confidence = 0.251813 | |
| I0330 12:18:29.876754 13762 sgd_solver.cpp:106] Iteration 24000, lr = 0.01 | |
| I0330 12:20:39.660506 13762 solver.cpp:229] Iteration 24500, loss = 2.66524 | |
| I0330 12:20:39.660794 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.315789 | |
| I0330 12:20:39.660825 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 12:20:39.660838 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.631579 | |
| I0330 12:20:39.660856 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.30922 (* 0.3 = 0.692765 loss) | |
| I0330 12:20:39.660871 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.637116 (* 0.3 = 0.191135 loss) | |
| I0330 12:20:39.660883 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.447368 | |
| I0330 12:20:39.660897 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 12:20:39.660909 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.684211 | |
| I0330 12:20:39.660923 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.86597 (* 0.3 = 0.55979 loss) | |
| I0330 12:20:39.660938 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.539818 (* 0.3 = 0.161945 loss) | |
| I0330 12:20:39.660950 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.578947 | |
| I0330 12:20:39.660962 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 12:20:39.660975 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.710526 | |
| I0330 12:20:39.660989 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3337 (* 1 = 1.3337 loss) | |
| I0330 12:20:39.661003 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.349691 (* 1 = 0.349691 loss) | |
| I0330 12:20:39.661016 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 12:20:39.661028 13762 solver.cpp:245] Train net output #16: total_confidence = 0.168151 | |
| I0330 12:20:39.661041 13762 sgd_solver.cpp:106] Iteration 24500, lr = 0.01 | |
| I0330 12:22:49.216577 13762 solver.cpp:338] Iteration 25000, Testing net (#0) | |
| I0330 12:23:19.865907 13762 solver.cpp:393] Test loss: 2.5056 | |
| I0330 12:23:19.866041 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.473149 | |
| I0330 12:23:19.866061 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.863276 | |
| I0330 12:23:19.866075 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.75322 | |
| I0330 12:23:19.866091 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.83853 (* 0.3 = 0.551558 loss) | |
| I0330 12:23:19.866106 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.483958 (* 0.3 = 0.145187 loss) | |
| I0330 12:23:19.866119 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.627988 | |
| I0330 12:23:19.866132 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.904957 | |
| I0330 12:23:19.866143 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.837244 | |
| I0330 12:23:19.866158 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.37536 (* 0.3 = 0.412609 loss) | |
| I0330 12:23:19.866174 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.359654 (* 0.3 = 0.107896 loss) | |
| I0330 12:23:19.866186 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.756659 | |
| I0330 12:23:19.866199 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.938683 | |
| I0330 12:23:19.866210 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.874143 | |
| I0330 12:23:19.866225 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.02571 (* 1 = 1.02571 loss) | |
| I0330 12:23:19.866240 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.262641 (* 1 = 0.262641 loss) | |
| I0330 12:23:19.866251 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.426 | |
| I0330 12:23:19.866262 13762 solver.cpp:406] Test net output #16: total_confidence = 0.384063 | |
| I0330 12:23:20.018807 13762 solver.cpp:229] Iteration 25000, loss = 2.77914 | |
| I0330 12:23:20.018885 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0330 12:23:20.018903 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 12:23:20.018916 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.586957 | |
| I0330 12:23:20.018934 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.16556 (* 0.3 = 0.649668 loss) | |
| I0330 12:23:20.018949 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.62031 (* 0.3 = 0.186093 loss) | |
| I0330 12:23:20.018961 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.543478 | |
| I0330 12:23:20.018986 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 12:23:20.019002 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.804348 | |
| I0330 12:23:20.019017 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.56784 (* 0.3 = 0.470353 loss) | |
| I0330 12:23:20.019032 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.461081 (* 0.3 = 0.138324 loss) | |
| I0330 12:23:20.019045 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.869565 | |
| I0330 12:23:20.019058 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 12:23:20.019070 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0330 12:23:20.019086 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.564446 (* 1 = 0.564446 loss) | |
| I0330 12:23:20.019100 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.171308 (* 1 = 0.171308 loss) | |
| I0330 12:23:20.019114 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 12:23:20.019125 13762 solver.cpp:245] Train net output #16: total_confidence = 0.338846 | |
| I0330 12:23:20.019139 13762 sgd_solver.cpp:106] Iteration 25000, lr = 0.01 | |
| I0330 12:25:29.621228 13762 solver.cpp:229] Iteration 25500, loss = 2.7712 | |
| I0330 12:25:29.621536 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.355556 | |
| I0330 12:25:29.621557 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 12:25:29.621572 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.555556 | |
| I0330 12:25:29.621588 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12631 (* 0.3 = 0.637893 loss) | |
| I0330 12:25:29.621603 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.638173 (* 0.3 = 0.191452 loss) | |
| I0330 12:25:29.621616 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377778 | |
| I0330 12:25:29.621628 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 12:25:29.621641 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.622222 | |
| I0330 12:25:29.621655 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9789 (* 0.3 = 0.59367 loss) | |
| I0330 12:25:29.621670 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.633777 (* 0.3 = 0.190133 loss) | |
| I0330 12:25:29.621683 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.577778 | |
| I0330 12:25:29.621696 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 12:25:29.621708 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0330 12:25:29.621723 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49056 (* 1 = 1.49056 loss) | |
| I0330 12:25:29.621737 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.545186 (* 1 = 0.545186 loss) | |
| I0330 12:25:29.621750 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 12:25:29.621762 13762 solver.cpp:245] Train net output #16: total_confidence = 0.260187 | |
| I0330 12:25:29.621775 13762 sgd_solver.cpp:106] Iteration 25500, lr = 0.01 | |
| I0330 12:27:39.324249 13762 solver.cpp:229] Iteration 26000, loss = 2.77272 | |
| I0330 12:27:39.324389 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.446809 | |
| I0330 12:27:39.324409 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 12:27:39.324422 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.702128 | |
| I0330 12:27:39.324439 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.91823 (* 0.3 = 0.575468 loss) | |
| I0330 12:27:39.324460 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.594969 (* 0.3 = 0.178491 loss) | |
| I0330 12:27:39.324473 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.595745 | |
| I0330 12:27:39.324486 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 12:27:39.324499 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.765957 | |
| I0330 12:27:39.324513 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.49042 (* 0.3 = 0.447127 loss) | |
| I0330 12:27:39.324527 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.493519 (* 0.3 = 0.148056 loss) | |
| I0330 12:27:39.324540 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829787 | |
| I0330 12:27:39.324559 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 12:27:39.324578 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0330 12:27:39.324592 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.790653 (* 1 = 0.790653 loss) | |
| I0330 12:27:39.324606 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.259803 (* 1 = 0.259803 loss) | |
| I0330 12:27:39.324618 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 12:27:39.324635 13762 solver.cpp:245] Train net output #16: total_confidence = 0.488719 | |
| I0330 12:27:39.324646 13762 sgd_solver.cpp:106] Iteration 26000, lr = 0.01 | |
| I0330 12:29:48.286525 13762 solver.cpp:229] Iteration 26500, loss = 2.76596 | |
| I0330 12:29:48.286633 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488889 | |
| I0330 12:29:48.286651 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 12:29:48.286664 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.755556 | |
| I0330 12:29:48.286680 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.90261 (* 0.3 = 0.570784 loss) | |
| I0330 12:29:48.286695 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.562878 (* 0.3 = 0.168863 loss) | |
| I0330 12:29:48.286707 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.533333 | |
| I0330 12:29:48.286723 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 12:29:48.286736 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.822222 | |
| I0330 12:29:48.286751 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.43737 (* 0.3 = 0.431211 loss) | |
| I0330 12:29:48.286766 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.428263 (* 0.3 = 0.128479 loss) | |
| I0330 12:29:48.286777 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0330 12:29:48.286790 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 12:29:48.286803 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.866667 | |
| I0330 12:29:48.286824 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.737295 (* 1 = 0.737295 loss) | |
| I0330 12:29:48.286839 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.206111 (* 1 = 0.206111 loss) | |
| I0330 12:29:48.286851 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:29:48.286864 13762 solver.cpp:245] Train net output #16: total_confidence = 0.328839 | |
| I0330 12:29:48.286875 13762 sgd_solver.cpp:106] Iteration 26500, lr = 0.01 | |
| I0330 12:31:56.982998 13762 solver.cpp:229] Iteration 27000, loss = 2.74364 | |
| I0330 12:31:56.983163 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 12:31:56.983183 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 12:31:56.983196 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 12:31:56.983220 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11561 (* 0.3 = 0.634682 loss) | |
| I0330 12:31:56.983235 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.630755 (* 0.3 = 0.189226 loss) | |
| I0330 12:31:56.983248 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0330 12:31:56.983260 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 12:31:56.983273 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.68 | |
| I0330 12:31:56.983288 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.87326 (* 0.3 = 0.561977 loss) | |
| I0330 12:31:56.983301 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.564415 (* 0.3 = 0.169325 loss) | |
| I0330 12:31:56.983314 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.68 | |
| I0330 12:31:56.983326 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 12:31:56.983338 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76 | |
| I0330 12:31:56.983352 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14043 (* 1 = 1.14043 loss) | |
| I0330 12:31:56.983367 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.365241 (* 1 = 0.365241 loss) | |
| I0330 12:31:56.983379 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:31:56.983392 13762 solver.cpp:245] Train net output #16: total_confidence = 0.245061 | |
| I0330 12:31:56.983403 13762 sgd_solver.cpp:106] Iteration 27000, lr = 0.01 | |
| I0330 12:34:05.891736 13762 solver.cpp:229] Iteration 27500, loss = 2.72533 | |
| I0330 12:34:05.891860 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.320755 | |
| I0330 12:34:05.891881 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 12:34:05.891894 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54717 | |
| I0330 12:34:05.891911 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34197 (* 0.3 = 0.702592 loss) | |
| I0330 12:34:05.891926 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.711954 (* 0.3 = 0.213586 loss) | |
| I0330 12:34:05.891937 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.339623 | |
| I0330 12:34:05.891950 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 12:34:05.891963 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.698113 | |
| I0330 12:34:05.891978 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.28468 (* 0.3 = 0.685405 loss) | |
| I0330 12:34:05.891991 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.698338 (* 0.3 = 0.209501 loss) | |
| I0330 12:34:05.892004 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.679245 | |
| I0330 12:34:05.892016 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 12:34:05.892030 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.811321 | |
| I0330 12:34:05.892043 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.21524 (* 1 = 1.21524 loss) | |
| I0330 12:34:05.892057 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.371698 (* 1 = 0.371698 loss) | |
| I0330 12:34:05.892069 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:34:05.892097 13762 solver.cpp:245] Train net output #16: total_confidence = 0.319224 | |
| I0330 12:34:05.892110 13762 sgd_solver.cpp:106] Iteration 27500, lr = 0.01 | |
| I0330 12:36:14.684433 13762 solver.cpp:229] Iteration 28000, loss = 2.68303 | |
| I0330 12:36:14.684564 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.38 | |
| I0330 12:36:14.684586 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 12:36:14.684598 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.7 | |
| I0330 12:36:14.684615 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.88417 (* 0.3 = 0.56525 loss) | |
| I0330 12:36:14.684630 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.62011 (* 0.3 = 0.186033 loss) | |
| I0330 12:36:14.684643 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.56 | |
| I0330 12:36:14.684655 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 12:36:14.684669 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76 | |
| I0330 12:36:14.684682 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.597 (* 0.3 = 0.4791 loss) | |
| I0330 12:36:14.684697 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524407 (* 0.3 = 0.157322 loss) | |
| I0330 12:36:14.684710 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.74 | |
| I0330 12:36:14.684721 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 12:36:14.684736 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0330 12:36:14.684751 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.815732 (* 1 = 0.815732 loss) | |
| I0330 12:36:14.684765 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.247697 (* 1 = 0.247697 loss) | |
| I0330 12:36:14.684778 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:36:14.684790 13762 solver.cpp:245] Train net output #16: total_confidence = 0.391626 | |
| I0330 12:36:14.684803 13762 sgd_solver.cpp:106] Iteration 28000, lr = 0.01 | |
| I0330 12:38:23.708706 13762 solver.cpp:229] Iteration 28500, loss = 2.70079 | |
| I0330 12:38:23.708827 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.38 | |
| I0330 12:38:23.708845 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 12:38:23.708859 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.62 | |
| I0330 12:38:23.708876 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32644 (* 0.3 = 0.697933 loss) | |
| I0330 12:38:23.708891 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.728259 (* 0.3 = 0.218478 loss) | |
| I0330 12:38:23.708904 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0330 12:38:23.708916 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 12:38:23.708928 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.7 | |
| I0330 12:38:23.708943 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95152 (* 0.3 = 0.585456 loss) | |
| I0330 12:38:23.708957 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.605971 (* 0.3 = 0.181791 loss) | |
| I0330 12:38:23.708971 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 12:38:23.708997 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0330 12:38:23.709009 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72 | |
| I0330 12:38:23.709023 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82496 (* 1 = 1.82496 loss) | |
| I0330 12:38:23.709038 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.539432 (* 1 = 0.539432 loss) | |
| I0330 12:38:23.709056 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:38:23.709069 13762 solver.cpp:245] Train net output #16: total_confidence = 0.351032 | |
| I0330 12:38:23.709080 13762 sgd_solver.cpp:106] Iteration 28500, lr = 0.01 | |
| I0330 12:40:32.578465 13762 solver.cpp:229] Iteration 29000, loss = 2.77253 | |
| I0330 12:40:32.578613 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.409091 | |
| I0330 12:40:32.578634 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 12:40:32.578646 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0330 12:40:32.578663 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.65611 (* 0.3 = 0.496834 loss) | |
| I0330 12:40:32.578686 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.475296 (* 0.3 = 0.142589 loss) | |
| I0330 12:40:32.578699 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431818 | |
| I0330 12:40:32.578712 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 12:40:32.578724 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.818182 | |
| I0330 12:40:32.578738 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48869 (* 0.3 = 0.446607 loss) | |
| I0330 12:40:32.578753 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.392974 (* 0.3 = 0.117892 loss) | |
| I0330 12:40:32.578766 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.704545 | |
| I0330 12:40:32.578778 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 12:40:32.578791 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0330 12:40:32.578805 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.865588 (* 1 = 0.865588 loss) | |
| I0330 12:40:32.578819 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.235679 (* 1 = 0.235679 loss) | |
| I0330 12:40:32.578833 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 12:40:32.578845 13762 solver.cpp:245] Train net output #16: total_confidence = 0.26838 | |
| I0330 12:40:32.578857 13762 sgd_solver.cpp:106] Iteration 29000, lr = 0.01 | |
| I0330 12:42:41.238932 13762 solver.cpp:229] Iteration 29500, loss = 2.81785 | |
| I0330 12:42:41.239042 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0330 12:42:41.239063 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 12:42:41.239075 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0330 12:42:41.239091 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.35077 (* 0.3 = 0.705232 loss) | |
| I0330 12:42:41.239106 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.737736 (* 0.3 = 0.221321 loss) | |
| I0330 12:42:41.239120 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.469388 | |
| I0330 12:42:41.239131 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 12:42:41.239145 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.693878 | |
| I0330 12:42:41.239158 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.87525 (* 0.3 = 0.562574 loss) | |
| I0330 12:42:41.239172 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.577039 (* 0.3 = 0.173112 loss) | |
| I0330 12:42:41.239186 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.632653 | |
| I0330 12:42:41.239197 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 12:42:41.239209 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0330 12:42:41.239224 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.23265 (* 1 = 1.23265 loss) | |
| I0330 12:42:41.239238 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.395121 (* 1 = 0.395121 loss) | |
| I0330 12:42:41.239250 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 12:42:41.239262 13762 solver.cpp:245] Train net output #16: total_confidence = 0.128943 | |
| I0330 12:42:41.239275 13762 sgd_solver.cpp:106] Iteration 29500, lr = 0.01 | |
| I0330 12:44:49.802253 13762 solver.cpp:338] Iteration 30000, Testing net (#0) | |
| I0330 12:45:19.690891 13762 solver.cpp:393] Test loss: 2.34123 | |
| I0330 12:45:19.690948 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.503017 | |
| I0330 12:45:19.690990 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.865412 | |
| I0330 12:45:19.691018 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.781566 | |
| I0330 12:45:19.691046 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.66868 (* 0.3 = 0.500605 loss) | |
| I0330 12:45:19.691074 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.461952 (* 0.3 = 0.138585 loss) | |
| I0330 12:45:19.691099 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.65216 | |
| I0330 12:45:19.691123 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.909957 | |
| I0330 12:45:19.691144 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.856452 | |
| I0330 12:45:19.691174 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.26073 (* 0.3 = 0.37822 loss) | |
| I0330 12:45:19.691207 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.332788 (* 0.3 = 0.0998363 loss) | |
| I0330 12:45:19.691229 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.763295 | |
| I0330 12:45:19.691259 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.939728 | |
| I0330 12:45:19.691279 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.889261 | |
| I0330 12:45:19.691304 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.971984 (* 1 = 0.971984 loss) | |
| I0330 12:45:19.691330 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.252002 (* 1 = 0.252002 loss) | |
| I0330 12:45:19.691351 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.418 | |
| I0330 12:45:19.691372 13762 solver.cpp:406] Test net output #16: total_confidence = 0.393239 | |
| I0330 12:45:19.843601 13762 solver.cpp:229] Iteration 30000, loss = 2.70206 | |
| I0330 12:45:19.843705 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.454545 | |
| I0330 12:45:19.843735 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 12:45:19.843760 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0330 12:45:19.843788 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.81241 (* 0.3 = 0.543722 loss) | |
| I0330 12:45:19.843816 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.520623 (* 0.3 = 0.156187 loss) | |
| I0330 12:45:19.843842 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.568182 | |
| I0330 12:45:19.843868 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 12:45:19.843891 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.886364 | |
| I0330 12:45:19.843925 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.33636 (* 0.3 = 0.400907 loss) | |
| I0330 12:45:19.843950 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.406799 (* 0.3 = 0.12204 loss) | |
| I0330 12:45:19.843981 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.886364 | |
| I0330 12:45:19.844003 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0330 12:45:19.844024 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.954545 | |
| I0330 12:45:19.844050 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.423909 (* 1 = 0.423909 loss) | |
| I0330 12:45:19.844076 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.124952 (* 1 = 0.124952 loss) | |
| I0330 12:45:19.844099 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:45:19.844121 13762 solver.cpp:245] Train net output #16: total_confidence = 0.392127 | |
| I0330 12:45:19.844143 13762 sgd_solver.cpp:106] Iteration 30000, lr = 0.01 | |
| I0330 12:47:28.678735 13762 solver.cpp:229] Iteration 30500, loss = 2.73076 | |
| I0330 12:47:28.678876 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.354167 | |
| I0330 12:47:28.678897 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 12:47:28.678910 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.541667 | |
| I0330 12:47:28.678932 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17414 (* 0.3 = 0.652242 loss) | |
| I0330 12:47:28.678948 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.61517 (* 0.3 = 0.184551 loss) | |
| I0330 12:47:28.678961 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 12:47:28.678973 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 12:47:28.678985 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0330 12:47:28.679018 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67316 (* 0.3 = 0.501949 loss) | |
| I0330 12:47:28.679033 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.493941 (* 0.3 = 0.148182 loss) | |
| I0330 12:47:28.679046 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.791667 | |
| I0330 12:47:28.679059 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 12:47:28.679071 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.854167 | |
| I0330 12:47:28.679085 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.894188 (* 1 = 0.894188 loss) | |
| I0330 12:47:28.679108 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.266044 (* 1 = 0.266044 loss) | |
| I0330 12:47:28.679121 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 12:47:28.679133 13762 solver.cpp:245] Train net output #16: total_confidence = 0.293013 | |
| I0330 12:47:28.679146 13762 sgd_solver.cpp:106] Iteration 30500, lr = 0.01 | |
| I0330 12:49:37.344077 13762 solver.cpp:229] Iteration 31000, loss = 2.69331 | |
| I0330 12:49:37.344189 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.295455 | |
| I0330 12:49:37.344209 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 12:49:37.344221 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.545455 | |
| I0330 12:49:37.344238 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.31706 (* 0.3 = 0.695118 loss) | |
| I0330 12:49:37.344252 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.672796 (* 0.3 = 0.201839 loss) | |
| I0330 12:49:37.344264 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.386364 | |
| I0330 12:49:37.344277 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 12:49:37.344290 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.659091 | |
| I0330 12:49:37.344305 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.07283 (* 0.3 = 0.62185 loss) | |
| I0330 12:49:37.344318 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.568095 (* 0.3 = 0.170428 loss) | |
| I0330 12:49:37.344331 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0330 12:49:37.344342 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 12:49:37.344367 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.772727 | |
| I0330 12:49:37.344383 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.30031 (* 1 = 1.30031 loss) | |
| I0330 12:49:37.344398 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.352925 (* 1 = 0.352925 loss) | |
| I0330 12:49:37.344409 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:49:37.344429 13762 solver.cpp:245] Train net output #16: total_confidence = 0.318024 | |
| I0330 12:49:37.344440 13762 sgd_solver.cpp:106] Iteration 31000, lr = 0.01 | |
| I0330 12:51:46.171252 13762 solver.cpp:229] Iteration 31500, loss = 2.69689 | |
| I0330 12:51:46.171399 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.358491 | |
| I0330 12:51:46.171419 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 12:51:46.171432 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.716981 | |
| I0330 12:51:46.171457 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02753 (* 0.3 = 0.60826 loss) | |
| I0330 12:51:46.171483 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.660589 (* 0.3 = 0.198177 loss) | |
| I0330 12:51:46.171496 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54717 | |
| I0330 12:51:46.171509 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 12:51:46.171521 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.754717 | |
| I0330 12:51:46.171535 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.50142 (* 0.3 = 0.450426 loss) | |
| I0330 12:51:46.171550 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.47505 (* 0.3 = 0.142515 loss) | |
| I0330 12:51:46.171562 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.698113 | |
| I0330 12:51:46.171576 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 12:51:46.171586 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.849057 | |
| I0330 12:51:46.171604 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.0656 (* 1 = 1.0656 loss) | |
| I0330 12:51:46.171618 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.386873 (* 1 = 0.386873 loss) | |
| I0330 12:51:46.171630 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:51:46.171643 13762 solver.cpp:245] Train net output #16: total_confidence = 0.185462 | |
| I0330 12:51:46.171668 13762 sgd_solver.cpp:106] Iteration 31500, lr = 0.01 | |
| I0330 12:53:54.847961 13762 solver.cpp:229] Iteration 32000, loss = 2.78786 | |
| I0330 12:53:54.848093 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145833 | |
| I0330 12:53:54.848112 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0330 12:53:54.848125 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.479167 | |
| I0330 12:53:54.848142 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68474 (* 0.3 = 0.805423 loss) | |
| I0330 12:53:54.848157 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.87859 (* 0.3 = 0.263577 loss) | |
| I0330 12:53:54.848172 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.354167 | |
| I0330 12:53:54.848186 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 12:53:54.848197 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.541667 | |
| I0330 12:53:54.848212 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.28115 (* 0.3 = 0.684344 loss) | |
| I0330 12:53:54.848227 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.733115 (* 0.3 = 0.219934 loss) | |
| I0330 12:53:54.848239 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.520833 | |
| I0330 12:53:54.848251 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0330 12:53:54.848263 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 12:53:54.848278 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.40483 (* 1 = 1.40483 loss) | |
| I0330 12:53:54.848292 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.500337 (* 1 = 0.500337 loss) | |
| I0330 12:53:54.848305 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:53:54.848317 13762 solver.cpp:245] Train net output #16: total_confidence = 0.177313 | |
| I0330 12:53:54.848330 13762 sgd_solver.cpp:106] Iteration 32000, lr = 0.01 | |
| I0330 12:56:03.603397 13762 solver.cpp:229] Iteration 32500, loss = 2.70605 | |
| I0330 12:56:03.603538 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 12:56:03.603559 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 12:56:03.603574 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5625 | |
| I0330 12:56:03.603605 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.61767 (* 0.3 = 0.7853 loss) | |
| I0330 12:56:03.603621 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.766115 (* 0.3 = 0.229835 loss) | |
| I0330 12:56:03.603634 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 12:56:03.603647 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 12:56:03.603659 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.708333 | |
| I0330 12:56:03.603673 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79999 (* 0.3 = 0.539998 loss) | |
| I0330 12:56:03.603688 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.562137 (* 0.3 = 0.168641 loss) | |
| I0330 12:56:03.603700 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.708333 | |
| I0330 12:56:03.603713 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 12:56:03.603725 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0330 12:56:03.603742 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.16355 (* 1 = 1.16355 loss) | |
| I0330 12:56:03.603757 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.341717 (* 1 = 0.341717 loss) | |
| I0330 12:56:03.603778 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:56:03.603798 13762 solver.cpp:245] Train net output #16: total_confidence = 0.191646 | |
| I0330 12:56:03.603811 13762 sgd_solver.cpp:106] Iteration 32500, lr = 0.01 | |
| I0330 12:58:12.294498 13762 solver.cpp:229] Iteration 33000, loss = 2.77281 | |
| I0330 12:58:12.294616 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.384615 | |
| I0330 12:58:12.294636 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 12:58:12.294648 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.653846 | |
| I0330 12:58:12.294666 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93629 (* 0.3 = 0.580887 loss) | |
| I0330 12:58:12.294679 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.591639 (* 0.3 = 0.177492 loss) | |
| I0330 12:58:12.294692 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.596154 | |
| I0330 12:58:12.294704 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 12:58:12.294716 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.865385 | |
| I0330 12:58:12.294730 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39656 (* 0.3 = 0.418968 loss) | |
| I0330 12:58:12.294744 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.43902 (* 0.3 = 0.131706 loss) | |
| I0330 12:58:12.294757 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.846154 | |
| I0330 12:58:12.294770 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 12:58:12.294781 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.942308 | |
| I0330 12:58:12.294795 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.552052 (* 1 = 0.552052 loss) | |
| I0330 12:58:12.294811 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.180088 (* 1 = 0.180088 loss) | |
| I0330 12:58:12.294823 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 12:58:12.294836 13762 solver.cpp:245] Train net output #16: total_confidence = 0.317472 | |
| I0330 12:58:12.294848 13762 sgd_solver.cpp:106] Iteration 33000, lr = 0.01 | |
| I0330 13:00:20.997781 13762 solver.cpp:229] Iteration 33500, loss = 2.69981 | |
| I0330 13:00:20.997937 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.294118 | |
| I0330 13:00:20.997959 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 13:00:20.997992 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509804 | |
| I0330 13:00:20.998010 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.31999 (* 0.3 = 0.695998 loss) | |
| I0330 13:00:20.998025 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.701113 (* 0.3 = 0.210334 loss) | |
| I0330 13:00:20.998039 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.470588 | |
| I0330 13:00:20.998051 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 13:00:20.998064 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72549 | |
| I0330 13:00:20.998077 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66224 (* 0.3 = 0.498673 loss) | |
| I0330 13:00:20.998092 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.528528 (* 0.3 = 0.158558 loss) | |
| I0330 13:00:20.998106 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.823529 | |
| I0330 13:00:20.998117 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 13:00:20.998136 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.882353 | |
| I0330 13:00:20.998150 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.825355 (* 1 = 0.825355 loss) | |
| I0330 13:00:20.998177 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27132 (* 1 = 0.27132 loss) | |
| I0330 13:00:20.998199 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:00:20.998211 13762 solver.cpp:245] Train net output #16: total_confidence = 0.226737 | |
| I0330 13:00:20.998224 13762 sgd_solver.cpp:106] Iteration 33500, lr = 0.01 | |
| I0330 13:02:29.721951 13762 solver.cpp:229] Iteration 34000, loss = 2.70384 | |
| I0330 13:02:29.722059 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4375 | |
| I0330 13:02:29.722077 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 13:02:29.722090 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 13:02:29.722106 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.95581 (* 0.3 = 0.586744 loss) | |
| I0330 13:02:29.722121 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559686 (* 0.3 = 0.167906 loss) | |
| I0330 13:02:29.722134 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5625 | |
| I0330 13:02:29.722147 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 13:02:29.722163 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8125 | |
| I0330 13:02:29.722177 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.45483 (* 0.3 = 0.436449 loss) | |
| I0330 13:02:29.722193 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.446645 (* 0.3 = 0.133994 loss) | |
| I0330 13:02:29.722205 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.791667 | |
| I0330 13:02:29.722218 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 13:02:29.722231 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 13:02:29.722245 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.81163 (* 1 = 0.81163 loss) | |
| I0330 13:02:29.722260 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.256415 (* 1 = 0.256415 loss) | |
| I0330 13:02:29.722273 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 13:02:29.722290 13762 solver.cpp:245] Train net output #16: total_confidence = 0.495643 | |
| I0330 13:02:29.722301 13762 sgd_solver.cpp:106] Iteration 34000, lr = 0.01 | |
| I0330 13:04:38.388589 13762 solver.cpp:229] Iteration 34500, loss = 2.72413 | |
| I0330 13:04:38.388736 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0330 13:04:38.388756 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 13:04:38.388769 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0330 13:04:38.388785 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65588 (* 0.3 = 0.796765 loss) | |
| I0330 13:04:38.388800 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.806723 (* 0.3 = 0.242017 loss) | |
| I0330 13:04:38.388813 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.520833 | |
| I0330 13:04:38.388825 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 13:04:38.388838 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.708333 | |
| I0330 13:04:38.388851 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.47409 (* 0.3 = 0.742228 loss) | |
| I0330 13:04:38.388866 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.77424 (* 0.3 = 0.232272 loss) | |
| I0330 13:04:38.388878 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.604167 | |
| I0330 13:04:38.388890 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 13:04:38.388902 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.770833 | |
| I0330 13:04:38.388916 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.36627 (* 1 = 2.36627 loss) | |
| I0330 13:04:38.388931 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.690877 (* 1 = 0.690877 loss) | |
| I0330 13:04:38.388943 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 13:04:38.388955 13762 solver.cpp:245] Train net output #16: total_confidence = 0.241029 | |
| I0330 13:04:38.388968 13762 sgd_solver.cpp:106] Iteration 34500, lr = 0.01 | |
| I0330 13:06:47.096328 13762 solver.cpp:338] Iteration 35000, Testing net (#0) | |
| I0330 13:07:17.019393 13762 solver.cpp:393] Test loss: 2.38865 | |
| I0330 13:07:17.019439 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.492086 | |
| I0330 13:07:17.019455 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.867184 | |
| I0330 13:07:17.019469 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.769351 | |
| I0330 13:07:17.019484 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.74736 (* 0.3 = 0.524208 loss) | |
| I0330 13:07:17.019498 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.462847 (* 0.3 = 0.138854 loss) | |
| I0330 13:07:17.019511 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.620808 | |
| I0330 13:07:17.019523 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.904139 | |
| I0330 13:07:17.019536 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.859216 | |
| I0330 13:07:17.019548 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.30591 (* 0.3 = 0.391773 loss) | |
| I0330 13:07:17.019563 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.337704 (* 0.3 = 0.101311 loss) | |
| I0330 13:07:17.019575 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.764495 | |
| I0330 13:07:17.019587 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.940092 | |
| I0330 13:07:17.019599 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.882507 | |
| I0330 13:07:17.019613 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.979334 (* 1 = 0.979334 loss) | |
| I0330 13:07:17.019628 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.253173 (* 1 = 0.253173 loss) | |
| I0330 13:07:17.019640 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.436 | |
| I0330 13:07:17.019652 13762 solver.cpp:406] Test net output #16: total_confidence = 0.375331 | |
| I0330 13:07:17.171092 13762 solver.cpp:229] Iteration 35000, loss = 2.69221 | |
| I0330 13:07:17.171218 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326087 | |
| I0330 13:07:17.171238 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 13:07:17.171252 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.630435 | |
| I0330 13:07:17.171275 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11983 (* 0.3 = 0.635949 loss) | |
| I0330 13:07:17.171289 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.604367 (* 0.3 = 0.18131 loss) | |
| I0330 13:07:17.171303 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.543478 | |
| I0330 13:07:17.171314 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 13:07:17.171327 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0330 13:07:17.171341 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61701 (* 0.3 = 0.485104 loss) | |
| I0330 13:07:17.171356 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.467368 (* 0.3 = 0.140211 loss) | |
| I0330 13:07:17.171368 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.804348 | |
| I0330 13:07:17.171380 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 13:07:17.171392 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.869565 | |
| I0330 13:07:17.171409 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.756027 (* 1 = 0.756027 loss) | |
| I0330 13:07:17.171423 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.236199 (* 1 = 0.236199 loss) | |
| I0330 13:07:17.171435 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:07:17.171447 13762 solver.cpp:245] Train net output #16: total_confidence = 0.433512 | |
| I0330 13:07:17.171469 13762 sgd_solver.cpp:106] Iteration 35000, lr = 0.01 | |
| I0330 13:09:25.757942 13762 solver.cpp:229] Iteration 35500, loss = 2.70474 | |
| I0330 13:09:25.758047 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0330 13:09:25.758067 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 13:09:25.758080 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.795455 | |
| I0330 13:09:25.758097 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.7005 (* 0.3 = 0.510151 loss) | |
| I0330 13:09:25.758112 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.580571 (* 0.3 = 0.174171 loss) | |
| I0330 13:09:25.758126 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.659091 | |
| I0330 13:09:25.758138 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 13:09:25.758150 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0330 13:09:25.758167 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.1331 (* 0.3 = 0.339929 loss) | |
| I0330 13:09:25.758183 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.389596 (* 0.3 = 0.116879 loss) | |
| I0330 13:09:25.758196 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.909091 | |
| I0330 13:09:25.758209 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0330 13:09:25.758221 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.954545 | |
| I0330 13:09:25.758244 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.400842 (* 1 = 0.400842 loss) | |
| I0330 13:09:25.758258 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.140637 (* 1 = 0.140637 loss) | |
| I0330 13:09:25.758270 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:09:25.758282 13762 solver.cpp:245] Train net output #16: total_confidence = 0.317479 | |
| I0330 13:09:25.758303 13762 sgd_solver.cpp:106] Iteration 35500, lr = 0.01 | |
| I0330 13:11:34.455273 13762 solver.cpp:229] Iteration 36000, loss = 2.77488 | |
| I0330 13:11:34.455458 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.414634 | |
| I0330 13:11:34.455482 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 13:11:34.455494 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.658537 | |
| I0330 13:11:34.455518 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14103 (* 0.3 = 0.642308 loss) | |
| I0330 13:11:34.455533 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.62618 (* 0.3 = 0.187854 loss) | |
| I0330 13:11:34.455548 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.463415 | |
| I0330 13:11:34.455559 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 13:11:34.455572 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.756098 | |
| I0330 13:11:34.455586 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65249 (* 0.3 = 0.495746 loss) | |
| I0330 13:11:34.455601 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.500849 (* 0.3 = 0.150255 loss) | |
| I0330 13:11:34.455615 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829268 | |
| I0330 13:11:34.455627 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 13:11:34.455639 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.951219 | |
| I0330 13:11:34.455654 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.663038 (* 1 = 0.663038 loss) | |
| I0330 13:11:34.455668 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.183312 (* 1 = 0.183312 loss) | |
| I0330 13:11:34.455682 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:11:34.455693 13762 solver.cpp:245] Train net output #16: total_confidence = 0.349814 | |
| I0330 13:11:34.455713 13762 sgd_solver.cpp:106] Iteration 36000, lr = 0.01 | |
| I0330 13:13:43.123888 13762 solver.cpp:229] Iteration 36500, loss = 2.66049 | |
| I0330 13:13:43.124016 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0330 13:13:43.124034 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 13:13:43.124053 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 13:13:43.124070 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10355 (* 0.3 = 0.631064 loss) | |
| I0330 13:13:43.124085 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.671008 (* 0.3 = 0.201302 loss) | |
| I0330 13:13:43.124099 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.666667 | |
| I0330 13:13:43.124111 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 13:13:43.124125 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0330 13:13:43.124140 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.22101 (* 0.3 = 0.366303 loss) | |
| I0330 13:13:43.124155 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.410592 (* 0.3 = 0.123178 loss) | |
| I0330 13:13:43.124171 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.911111 | |
| I0330 13:13:43.124182 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 13:13:43.124196 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 13:13:43.124212 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.288885 (* 1 = 0.288885 loss) | |
| I0330 13:13:43.124227 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.116789 (* 1 = 0.116789 loss) | |
| I0330 13:13:43.124239 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:13:43.124251 13762 solver.cpp:245] Train net output #16: total_confidence = 0.353041 | |
| I0330 13:13:43.124264 13762 sgd_solver.cpp:106] Iteration 36500, lr = 0.01 | |
| I0330 13:15:51.931499 13762 solver.cpp:229] Iteration 37000, loss = 2.73449 | |
| I0330 13:15:51.931639 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.369565 | |
| I0330 13:15:51.931659 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 13:15:51.931675 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.630435 | |
| I0330 13:15:51.931710 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13099 (* 0.3 = 0.639298 loss) | |
| I0330 13:15:51.931727 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.636197 (* 0.3 = 0.190859 loss) | |
| I0330 13:15:51.931740 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.521739 | |
| I0330 13:15:51.931753 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 13:15:51.931766 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673913 | |
| I0330 13:15:51.931779 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.62164 (* 0.3 = 0.486491 loss) | |
| I0330 13:15:51.931794 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.51929 (* 0.3 = 0.155787 loss) | |
| I0330 13:15:51.931807 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.73913 | |
| I0330 13:15:51.931819 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 13:15:51.931833 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0330 13:15:51.931854 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.78042 (* 1 = 0.78042 loss) | |
| I0330 13:15:51.931869 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.221242 (* 1 = 0.221242 loss) | |
| I0330 13:15:51.931881 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 13:15:51.931893 13762 solver.cpp:245] Train net output #16: total_confidence = 0.445458 | |
| I0330 13:15:51.931917 13762 sgd_solver.cpp:106] Iteration 37000, lr = 0.01 | |
| I0330 13:18:00.582000 13762 solver.cpp:229] Iteration 37500, loss = 2.65787 | |
| I0330 13:18:00.582110 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488372 | |
| I0330 13:18:00.582129 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 13:18:00.582142 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744186 | |
| I0330 13:18:00.582160 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.66821 (* 0.3 = 0.500462 loss) | |
| I0330 13:18:00.582176 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.475904 (* 0.3 = 0.142771 loss) | |
| I0330 13:18:00.582188 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.581395 | |
| I0330 13:18:00.582201 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 13:18:00.582214 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.837209 | |
| I0330 13:18:00.582228 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.28687 (* 0.3 = 0.38606 loss) | |
| I0330 13:18:00.582242 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.373948 (* 0.3 = 0.112185 loss) | |
| I0330 13:18:00.582255 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.767442 | |
| I0330 13:18:00.582274 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 13:18:00.582288 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.930233 | |
| I0330 13:18:00.582303 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.729116 (* 1 = 0.729116 loss) | |
| I0330 13:18:00.582316 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.23195 (* 1 = 0.23195 loss) | |
| I0330 13:18:00.582329 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 13:18:00.582345 13762 solver.cpp:245] Train net output #16: total_confidence = 0.347293 | |
| I0330 13:18:00.582357 13762 sgd_solver.cpp:106] Iteration 37500, lr = 0.01 | |
| I0330 13:20:09.383397 13762 solver.cpp:229] Iteration 38000, loss = 2.623 | |
| I0330 13:20:09.383620 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4375 | |
| I0330 13:20:09.383641 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 13:20:09.383662 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.729167 | |
| I0330 13:20:09.383680 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73587 (* 0.3 = 0.520762 loss) | |
| I0330 13:20:09.383695 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.520258 (* 0.3 = 0.156077 loss) | |
| I0330 13:20:09.383708 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.666667 | |
| I0330 13:20:09.383720 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 13:20:09.383733 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.895833 | |
| I0330 13:20:09.383747 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.0432 (* 0.3 = 0.312959 loss) | |
| I0330 13:20:09.383762 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.321758 (* 0.3 = 0.0965274 loss) | |
| I0330 13:20:09.383775 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.875 | |
| I0330 13:20:09.383788 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0330 13:20:09.383800 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0330 13:20:09.383816 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.526238 (* 1 = 0.526238 loss) | |
| I0330 13:20:09.383831 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.14624 (* 1 = 0.14624 loss) | |
| I0330 13:20:09.383842 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 13:20:09.383854 13762 solver.cpp:245] Train net output #16: total_confidence = 0.440595 | |
| I0330 13:20:09.383867 13762 sgd_solver.cpp:106] Iteration 38000, lr = 0.01 | |
| I0330 13:22:18.170195 13762 solver.cpp:229] Iteration 38500, loss = 2.72215 | |
| I0330 13:22:18.170301 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.28 | |
| I0330 13:22:18.170320 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 13:22:18.170334 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0330 13:22:18.170351 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.0886 (* 0.3 = 0.626579 loss) | |
| I0330 13:22:18.170367 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.61646 (* 0.3 = 0.184938 loss) | |
| I0330 13:22:18.170379 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.48 | |
| I0330 13:22:18.170393 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 13:22:18.170406 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.66 | |
| I0330 13:22:18.170420 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.82956 (* 0.3 = 0.548869 loss) | |
| I0330 13:22:18.170435 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.550754 (* 0.3 = 0.165226 loss) | |
| I0330 13:22:18.170447 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.74 | |
| I0330 13:22:18.170459 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 13:22:18.170472 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0330 13:22:18.170486 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.07029 (* 1 = 1.07029 loss) | |
| I0330 13:22:18.170500 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.323314 (* 1 = 0.323314 loss) | |
| I0330 13:22:18.170512 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 13:22:18.170526 13762 solver.cpp:245] Train net output #16: total_confidence = 0.228726 | |
| I0330 13:22:18.170537 13762 sgd_solver.cpp:106] Iteration 38500, lr = 0.01 | |
| I0330 13:24:26.931521 13762 solver.cpp:229] Iteration 39000, loss = 2.68908 | |
| I0330 13:24:26.931649 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44 | |
| I0330 13:24:26.931670 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 13:24:26.931684 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.68 | |
| I0330 13:24:26.931699 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80312 (* 0.3 = 0.540935 loss) | |
| I0330 13:24:26.931715 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.543304 (* 0.3 = 0.162991 loss) | |
| I0330 13:24:26.931730 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54 | |
| I0330 13:24:26.931743 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 13:24:26.931756 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 13:24:26.931771 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47537 (* 0.3 = 0.44261 loss) | |
| I0330 13:24:26.931784 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.428904 (* 0.3 = 0.128671 loss) | |
| I0330 13:24:26.931797 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.76 | |
| I0330 13:24:26.931808 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 13:24:26.931820 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.96 | |
| I0330 13:24:26.931834 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.718519 (* 1 = 0.718519 loss) | |
| I0330 13:24:26.931849 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.211472 (* 1 = 0.211472 loss) | |
| I0330 13:24:26.931875 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:24:26.931887 13762 solver.cpp:245] Train net output #16: total_confidence = 0.516915 | |
| I0330 13:24:26.931900 13762 sgd_solver.cpp:106] Iteration 39000, lr = 0.01 | |
| I0330 13:26:35.622306 13762 solver.cpp:229] Iteration 39500, loss = 2.68348 | |
| I0330 13:26:35.622458 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3 | |
| I0330 13:26:35.622478 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 13:26:35.622493 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.45 | |
| I0330 13:26:35.622509 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.035 (* 0.3 = 0.910501 loss) | |
| I0330 13:26:35.622524 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.817198 (* 0.3 = 0.245159 loss) | |
| I0330 13:26:35.622536 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 13:26:35.622550 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 13:26:35.622561 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 13:26:35.622575 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.93925 (* 0.3 = 0.581776 loss) | |
| I0330 13:26:35.622591 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.547479 (* 0.3 = 0.164244 loss) | |
| I0330 13:26:35.622603 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 13:26:35.622616 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 13:26:35.622635 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.725 | |
| I0330 13:26:35.622659 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.86085 (* 1 = 1.86085 loss) | |
| I0330 13:26:35.622675 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.477821 (* 1 = 0.477821 loss) | |
| I0330 13:26:35.622687 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 13:26:35.622700 13762 solver.cpp:245] Train net output #16: total_confidence = 0.150112 | |
| I0330 13:26:35.622712 13762 sgd_solver.cpp:106] Iteration 39500, lr = 0.01 | |
| I0330 13:28:43.940819 13762 solver.cpp:338] Iteration 40000, Testing net (#0) | |
| I0330 13:29:13.826197 13762 solver.cpp:393] Test loss: 2.30499 | |
| I0330 13:29:13.826244 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.528694 | |
| I0330 13:29:13.826261 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.868731 | |
| I0330 13:29:13.826274 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.791276 | |
| I0330 13:29:13.826289 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.59482 (* 0.3 = 0.478446 loss) | |
| I0330 13:29:13.826304 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.446585 (* 0.3 = 0.133976 loss) | |
| I0330 13:29:13.826316 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.662425 | |
| I0330 13:29:13.826329 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.910638 | |
| I0330 13:29:13.826341 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.857916 | |
| I0330 13:29:13.826355 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.2081 (* 0.3 = 0.362429 loss) | |
| I0330 13:29:13.826370 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.325046 (* 0.3 = 0.0975138 loss) | |
| I0330 13:29:13.826382 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.758604 | |
| I0330 13:29:13.826401 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.937455 | |
| I0330 13:29:13.826416 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.887957 | |
| I0330 13:29:13.826428 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.974658 (* 1 = 0.974658 loss) | |
| I0330 13:29:13.826442 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.25797 (* 1 = 0.25797 loss) | |
| I0330 13:29:13.826454 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.428 | |
| I0330 13:29:13.826474 13762 solver.cpp:406] Test net output #16: total_confidence = 0.418687 | |
| I0330 13:29:13.977921 13762 solver.cpp:229] Iteration 40000, loss = 2.61323 | |
| I0330 13:29:13.978013 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404255 | |
| I0330 13:29:13.978031 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 13:29:13.978044 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.702128 | |
| I0330 13:29:13.978060 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.65146 (* 0.3 = 0.495438 loss) | |
| I0330 13:29:13.978073 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.492223 (* 0.3 = 0.147667 loss) | |
| I0330 13:29:13.978086 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.638298 | |
| I0330 13:29:13.978098 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 13:29:13.978111 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.87234 | |
| I0330 13:29:13.978126 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.23712 (* 0.3 = 0.371135 loss) | |
| I0330 13:29:13.978139 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.361425 (* 0.3 = 0.108427 loss) | |
| I0330 13:29:13.978152 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.893617 | |
| I0330 13:29:13.978170 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 13:29:13.978190 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.93617 | |
| I0330 13:29:13.978204 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.476804 (* 1 = 0.476804 loss) | |
| I0330 13:29:13.978219 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.146384 (* 1 = 0.146384 loss) | |
| I0330 13:29:13.978231 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:29:13.978243 13762 solver.cpp:245] Train net output #16: total_confidence = 0.303136 | |
| I0330 13:29:13.978255 13762 sgd_solver.cpp:106] Iteration 40000, lr = 0.01 | |
| I0330 13:31:22.731655 13762 solver.cpp:229] Iteration 40500, loss = 2.69637 | |
| I0330 13:31:22.731796 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.452381 | |
| I0330 13:31:22.731815 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 13:31:22.731829 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.761905 | |
| I0330 13:31:22.731854 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.61243 (* 0.3 = 0.48373 loss) | |
| I0330 13:31:22.731869 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.471662 (* 0.3 = 0.141499 loss) | |
| I0330 13:31:22.731881 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.619048 | |
| I0330 13:31:22.731894 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 13:31:22.731907 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.97619 | |
| I0330 13:31:22.731921 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.02418 (* 0.3 = 0.307255 loss) | |
| I0330 13:31:22.731935 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.385472 (* 0.3 = 0.115642 loss) | |
| I0330 13:31:22.731953 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.928571 | |
| I0330 13:31:22.731976 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0330 13:31:22.731988 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 13:31:22.732002 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.197473 (* 1 = 0.197473 loss) | |
| I0330 13:31:22.732017 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.053396 (* 1 = 0.053396 loss) | |
| I0330 13:31:22.732038 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 13:31:22.732049 13762 solver.cpp:245] Train net output #16: total_confidence = 0.481516 | |
| I0330 13:31:22.732061 13762 sgd_solver.cpp:106] Iteration 40500, lr = 0.01 | |
| I0330 13:33:31.414681 13762 solver.cpp:229] Iteration 41000, loss = 2.74987 | |
| I0330 13:33:31.414790 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.315789 | |
| I0330 13:33:31.414810 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 13:33:31.414832 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.578947 | |
| I0330 13:33:31.414849 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58999 (* 0.3 = 0.776998 loss) | |
| I0330 13:33:31.414863 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.605784 (* 0.3 = 0.181735 loss) | |
| I0330 13:33:31.414875 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.473684 | |
| I0330 13:33:31.414894 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 13:33:31.414907 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.684211 | |
| I0330 13:33:31.414921 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.17896 (* 0.3 = 0.653689 loss) | |
| I0330 13:33:31.414935 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.540243 (* 0.3 = 0.162073 loss) | |
| I0330 13:33:31.414947 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.578947 | |
| I0330 13:33:31.414960 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 13:33:31.414983 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.736842 | |
| I0330 13:33:31.415016 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.71161 (* 1 = 1.71161 loss) | |
| I0330 13:33:31.415031 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.38773 (* 1 = 0.38773 loss) | |
| I0330 13:33:31.415045 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 13:33:31.415056 13762 solver.cpp:245] Train net output #16: total_confidence = 0.169338 | |
| I0330 13:33:31.415073 13762 sgd_solver.cpp:106] Iteration 41000, lr = 0.01 | |
| I0330 13:35:40.430583 13762 solver.cpp:229] Iteration 41500, loss = 2.70719 | |
| I0330 13:35:40.430732 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 13:35:40.430753 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 13:35:40.430766 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627451 | |
| I0330 13:35:40.430781 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.05937 (* 0.3 = 0.617811 loss) | |
| I0330 13:35:40.430796 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.623742 (* 0.3 = 0.187123 loss) | |
| I0330 13:35:40.430809 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.529412 | |
| I0330 13:35:40.430831 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 13:35:40.430843 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.784314 | |
| I0330 13:35:40.430857 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.5496 (* 0.3 = 0.464879 loss) | |
| I0330 13:35:40.430872 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.468763 (* 0.3 = 0.140629 loss) | |
| I0330 13:35:40.430886 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.705882 | |
| I0330 13:35:40.430897 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 13:35:40.430909 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.921569 | |
| I0330 13:35:40.430924 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.88913 (* 1 = 0.88913 loss) | |
| I0330 13:35:40.430938 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.26558 (* 1 = 0.26558 loss) | |
| I0330 13:35:40.430950 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 13:35:40.430963 13762 solver.cpp:245] Train net output #16: total_confidence = 0.240187 | |
| I0330 13:35:40.430990 13762 sgd_solver.cpp:106] Iteration 41500, lr = 0.01 | |
| I0330 13:37:49.283622 13762 solver.cpp:229] Iteration 42000, loss = 2.62424 | |
| I0330 13:37:49.283716 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.365385 | |
| I0330 13:37:49.283738 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 13:37:49.283751 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.615385 | |
| I0330 13:37:49.283767 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.24306 (* 0.3 = 0.672917 loss) | |
| I0330 13:37:49.283782 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.7069 (* 0.3 = 0.21207 loss) | |
| I0330 13:37:49.283795 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.384615 | |
| I0330 13:37:49.283807 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 13:37:49.283820 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 13:37:49.283834 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75718 (* 0.3 = 0.527153 loss) | |
| I0330 13:37:49.283849 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.545106 (* 0.3 = 0.163532 loss) | |
| I0330 13:37:49.283861 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 13:37:49.283874 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 13:37:49.283885 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0330 13:37:49.283911 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.949003 (* 1 = 0.949003 loss) | |
| I0330 13:37:49.283941 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.286757 (* 1 = 0.286757 loss) | |
| I0330 13:37:49.283965 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:37:49.283978 13762 solver.cpp:245] Train net output #16: total_confidence = 0.299101 | |
| I0330 13:37:49.283992 13762 sgd_solver.cpp:106] Iteration 42000, lr = 0.01 | |
| I0330 13:39:58.023334 13762 solver.cpp:229] Iteration 42500, loss = 2.62652 | |
| I0330 13:39:58.023483 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.339623 | |
| I0330 13:39:58.023504 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 13:39:58.023516 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.603774 | |
| I0330 13:39:58.023533 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.20253 (* 0.3 = 0.660758 loss) | |
| I0330 13:39:58.023548 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.696559 (* 0.3 = 0.208968 loss) | |
| I0330 13:39:58.023568 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45283 | |
| I0330 13:39:58.023581 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 13:39:58.023593 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.716981 | |
| I0330 13:39:58.023608 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.94125 (* 0.3 = 0.582376 loss) | |
| I0330 13:39:58.023622 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.601056 (* 0.3 = 0.180317 loss) | |
| I0330 13:39:58.023634 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.660377 | |
| I0330 13:39:58.023648 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 13:39:58.023659 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.90566 | |
| I0330 13:39:58.023679 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24951 (* 1 = 1.24951 loss) | |
| I0330 13:39:58.023715 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.425208 (* 1 = 0.425208 loss) | |
| I0330 13:39:58.023739 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 13:39:58.023771 13762 solver.cpp:245] Train net output #16: total_confidence = 0.121887 | |
| I0330 13:39:58.023792 13762 sgd_solver.cpp:106] Iteration 42500, lr = 0.01 | |
| I0330 13:42:07.184691 13762 solver.cpp:229] Iteration 43000, loss = 2.6848 | |
| I0330 13:42:07.184810 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.236842 | |
| I0330 13:42:07.184831 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 13:42:07.184844 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.657895 | |
| I0330 13:42:07.184860 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13973 (* 0.3 = 0.641919 loss) | |
| I0330 13:42:07.184875 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.621648 (* 0.3 = 0.186494 loss) | |
| I0330 13:42:07.184888 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.473684 | |
| I0330 13:42:07.184901 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 13:42:07.184913 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.815789 | |
| I0330 13:42:07.184927 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.517 (* 0.3 = 0.455101 loss) | |
| I0330 13:42:07.184942 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.41296 (* 0.3 = 0.123888 loss) | |
| I0330 13:42:07.184955 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.736842 | |
| I0330 13:42:07.184967 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 13:42:07.184979 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.894737 | |
| I0330 13:42:07.184993 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.888797 (* 1 = 0.888797 loss) | |
| I0330 13:42:07.185012 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.229333 (* 1 = 0.229333 loss) | |
| I0330 13:42:07.185032 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 13:42:07.185046 13762 solver.cpp:245] Train net output #16: total_confidence = 0.304734 | |
| I0330 13:42:07.185060 13762 sgd_solver.cpp:106] Iteration 43000, lr = 0.01 | |
| I0330 13:44:16.524889 13762 solver.cpp:229] Iteration 43500, loss = 2.64421 | |
| I0330 13:44:16.524996 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431373 | |
| I0330 13:44:16.525015 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 13:44:16.525028 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.784314 | |
| I0330 13:44:16.525045 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73726 (* 0.3 = 0.521179 loss) | |
| I0330 13:44:16.525059 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.543413 (* 0.3 = 0.163024 loss) | |
| I0330 13:44:16.525073 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.647059 | |
| I0330 13:44:16.525084 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0330 13:44:16.525097 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.921569 | |
| I0330 13:44:16.525111 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.05685 (* 0.3 = 0.317054 loss) | |
| I0330 13:44:16.525126 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.32274 (* 0.3 = 0.0968219 loss) | |
| I0330 13:44:16.525138 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.941176 | |
| I0330 13:44:16.525156 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0330 13:44:16.525172 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 13:44:16.525187 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.233078 (* 1 = 0.233078 loss) | |
| I0330 13:44:16.525200 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0740436 (* 1 = 0.0740436 loss) | |
| I0330 13:44:16.525213 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:44:16.525225 13762 solver.cpp:245] Train net output #16: total_confidence = 0.371925 | |
| I0330 13:44:16.525238 13762 sgd_solver.cpp:106] Iteration 43500, lr = 0.01 | |
| I0330 13:46:25.209079 13762 solver.cpp:229] Iteration 44000, loss = 2.63846 | |
| I0330 13:46:25.209241 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326923 | |
| I0330 13:46:25.209277 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 13:46:25.209293 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.692308 | |
| I0330 13:46:25.209311 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04374 (* 0.3 = 0.613122 loss) | |
| I0330 13:46:25.209324 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.626258 (* 0.3 = 0.187877 loss) | |
| I0330 13:46:25.209337 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 13:46:25.209350 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 13:46:25.209362 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.807692 | |
| I0330 13:46:25.209377 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.59178 (* 0.3 = 0.477533 loss) | |
| I0330 13:46:25.209391 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.495962 (* 0.3 = 0.148789 loss) | |
| I0330 13:46:25.209403 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.807692 | |
| I0330 13:46:25.209416 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 13:46:25.209429 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.865385 | |
| I0330 13:46:25.209442 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.778612 (* 1 = 0.778612 loss) | |
| I0330 13:46:25.209456 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.267871 (* 1 = 0.267871 loss) | |
| I0330 13:46:25.209472 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 13:46:25.209486 13762 solver.cpp:245] Train net output #16: total_confidence = 0.328985 | |
| I0330 13:46:25.209497 13762 sgd_solver.cpp:106] Iteration 44000, lr = 0.01 | |
| I0330 13:48:34.607008 13762 solver.cpp:229] Iteration 44500, loss = 2.63433 | |
| I0330 13:48:34.607125 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.422222 | |
| I0330 13:48:34.607146 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 13:48:34.607161 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.622222 | |
| I0330 13:48:34.607178 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.00811 (* 0.3 = 0.602432 loss) | |
| I0330 13:48:34.607193 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.643316 (* 0.3 = 0.192995 loss) | |
| I0330 13:48:34.607206 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.533333 | |
| I0330 13:48:34.607219 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 13:48:34.607231 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0330 13:48:34.607245 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48795 (* 0.3 = 0.446386 loss) | |
| I0330 13:48:34.607260 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.456393 (* 0.3 = 0.136918 loss) | |
| I0330 13:48:34.607275 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0330 13:48:34.607311 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 13:48:34.607331 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0330 13:48:34.607347 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.600607 (* 1 = 0.600607 loss) | |
| I0330 13:48:34.607368 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.166368 (* 1 = 0.166368 loss) | |
| I0330 13:48:34.607380 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 13:48:34.607393 13762 solver.cpp:245] Train net output #16: total_confidence = 0.449538 | |
| I0330 13:48:34.607404 13762 sgd_solver.cpp:106] Iteration 44500, lr = 0.01 | |
| I0330 13:50:43.173805 13762 solver.cpp:338] Iteration 45000, Testing net (#0) | |
| I0330 13:51:13.132184 13762 solver.cpp:393] Test loss: 2.16746 | |
| I0330 13:51:13.132233 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.544336 | |
| I0330 13:51:13.132249 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.877957 | |
| I0330 13:51:13.132262 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.805201 | |
| I0330 13:51:13.132285 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.56638 (* 0.3 = 0.469915 loss) | |
| I0330 13:51:13.132299 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.428488 (* 0.3 = 0.128546 loss) | |
| I0330 13:51:13.132311 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.681389 | |
| I0330 13:51:13.132323 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.915366 | |
| I0330 13:51:13.132336 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.870054 | |
| I0330 13:51:13.132350 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.1613 (* 0.3 = 0.348391 loss) | |
| I0330 13:51:13.132364 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.310668 (* 0.3 = 0.0932004 loss) | |
| I0330 13:51:13.132376 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.783986 | |
| I0330 13:51:13.132388 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.945183 | |
| I0330 13:51:13.132400 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.894659 | |
| I0330 13:51:13.132414 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.893794 (* 1 = 0.893794 loss) | |
| I0330 13:51:13.132431 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.23361 (* 1 = 0.23361 loss) | |
| I0330 13:51:13.132462 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.476 | |
| I0330 13:51:13.132493 13762 solver.cpp:406] Test net output #16: total_confidence = 0.40639 | |
| I0330 13:51:13.284112 13762 solver.cpp:229] Iteration 45000, loss = 2.579 | |
| I0330 13:51:13.284216 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 13:51:13.284235 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 13:51:13.284248 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.690476 | |
| I0330 13:51:13.284265 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12363 (* 0.3 = 0.637089 loss) | |
| I0330 13:51:13.284279 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.572835 (* 0.3 = 0.171851 loss) | |
| I0330 13:51:13.284291 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 13:51:13.284304 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 13:51:13.284317 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.809524 | |
| I0330 13:51:13.284330 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.55754 (* 0.3 = 0.467261 loss) | |
| I0330 13:51:13.284344 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.41392 (* 0.3 = 0.124176 loss) | |
| I0330 13:51:13.284358 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.809524 | |
| I0330 13:51:13.284369 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 13:51:13.284381 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97619 | |
| I0330 13:51:13.284396 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.557109 (* 1 = 0.557109 loss) | |
| I0330 13:51:13.284410 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.152305 (* 1 = 0.152305 loss) | |
| I0330 13:51:13.284422 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 13:51:13.284440 13762 solver.cpp:245] Train net output #16: total_confidence = 0.334556 | |
| I0330 13:51:13.284453 13762 sgd_solver.cpp:106] Iteration 45000, lr = 0.01 | |
| I0330 13:53:22.724480 13762 solver.cpp:229] Iteration 45500, loss = 2.63318 | |
| I0330 13:53:22.724642 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.444444 | |
| I0330 13:53:22.724663 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 13:53:22.724676 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 13:53:22.724699 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79395 (* 0.3 = 0.538184 loss) | |
| I0330 13:53:22.724714 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.523817 (* 0.3 = 0.157145 loss) | |
| I0330 13:53:22.724727 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.555556 | |
| I0330 13:53:22.724740 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 13:53:22.724751 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.822222 | |
| I0330 13:53:22.724766 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.22231 (* 0.3 = 0.366694 loss) | |
| I0330 13:53:22.724781 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.378033 (* 0.3 = 0.11341 loss) | |
| I0330 13:53:22.724793 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0330 13:53:22.724805 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 13:53:22.724817 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.955556 | |
| I0330 13:53:22.724838 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.52446 (* 1 = 0.52446 loss) | |
| I0330 13:53:22.724853 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.193123 (* 1 = 0.193123 loss) | |
| I0330 13:53:22.724865 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 13:53:22.724876 13762 solver.cpp:245] Train net output #16: total_confidence = 0.316622 | |
| I0330 13:53:22.724889 13762 sgd_solver.cpp:106] Iteration 45500, lr = 0.01 | |
| I0330 13:55:31.595819 13762 solver.cpp:229] Iteration 46000, loss = 2.60822 | |
| I0330 13:55:31.595944 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0330 13:55:31.595964 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 13:55:31.595978 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489796 | |
| I0330 13:55:31.595993 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.45563 (* 0.3 = 0.736689 loss) | |
| I0330 13:55:31.596009 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.730547 (* 0.3 = 0.219164 loss) | |
| I0330 13:55:31.596021 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510204 | |
| I0330 13:55:31.596035 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 13:55:31.596046 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0330 13:55:31.596061 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66278 (* 0.3 = 0.498835 loss) | |
| I0330 13:55:31.596076 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.496612 (* 0.3 = 0.148983 loss) | |
| I0330 13:55:31.596088 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0330 13:55:31.596101 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 13:55:31.596112 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0330 13:55:31.596127 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.12935 (* 1 = 1.12935 loss) | |
| I0330 13:55:31.596141 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.338928 (* 1 = 0.338928 loss) | |
| I0330 13:55:31.596153 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 13:55:31.596169 13762 solver.cpp:245] Train net output #16: total_confidence = 0.305896 | |
| I0330 13:55:31.596190 13762 sgd_solver.cpp:106] Iteration 46000, lr = 0.01 | |
| I0330 13:57:40.328487 13762 solver.cpp:229] Iteration 46500, loss = 2.63858 | |
| I0330 13:57:40.328649 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3125 | |
| I0330 13:57:40.328670 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 13:57:40.328683 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0330 13:57:40.328707 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.18449 (* 0.3 = 0.655346 loss) | |
| I0330 13:57:40.328722 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.644574 (* 0.3 = 0.193372 loss) | |
| I0330 13:57:40.328735 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.520833 | |
| I0330 13:57:40.328748 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 13:57:40.328760 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.729167 | |
| I0330 13:57:40.328774 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83565 (* 0.3 = 0.550695 loss) | |
| I0330 13:57:40.328789 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.520384 (* 0.3 = 0.156115 loss) | |
| I0330 13:57:40.328801 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.729167 | |
| I0330 13:57:40.328814 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 13:57:40.328826 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 13:57:40.328840 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.19553 (* 1 = 1.19553 loss) | |
| I0330 13:57:40.328857 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.369801 (* 1 = 0.369801 loss) | |
| I0330 13:57:40.328886 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 13:57:40.328909 13762 solver.cpp:245] Train net output #16: total_confidence = 0.198739 | |
| I0330 13:57:40.328933 13762 sgd_solver.cpp:106] Iteration 46500, lr = 0.01 | |
| I0330 13:59:49.124056 13762 solver.cpp:229] Iteration 47000, loss = 2.63105 | |
| I0330 13:59:49.124157 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.390244 | |
| I0330 13:59:49.124176 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 13:59:49.124189 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.658537 | |
| I0330 13:59:49.124205 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09183 (* 0.3 = 0.627549 loss) | |
| I0330 13:59:49.124219 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.583821 (* 0.3 = 0.175146 loss) | |
| I0330 13:59:49.124231 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.487805 | |
| I0330 13:59:49.124244 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 13:59:49.124256 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.804878 | |
| I0330 13:59:49.124271 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.57006 (* 0.3 = 0.471018 loss) | |
| I0330 13:59:49.124285 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.448683 (* 0.3 = 0.134605 loss) | |
| I0330 13:59:49.124297 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.780488 | |
| I0330 13:59:49.124310 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 13:59:49.124321 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.902439 | |
| I0330 13:59:49.124336 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.757205 (* 1 = 0.757205 loss) | |
| I0330 13:59:49.124366 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.224504 (* 1 = 0.224504 loss) | |
| I0330 13:59:49.124378 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 13:59:49.124390 13762 solver.cpp:245] Train net output #16: total_confidence = 0.27409 | |
| I0330 13:59:49.124402 13762 sgd_solver.cpp:106] Iteration 47000, lr = 0.01 | |
| I0330 14:01:58.030249 13762 solver.cpp:229] Iteration 47500, loss = 2.64615 | |
| I0330 14:01:58.030377 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 14:01:58.030397 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 14:01:58.030411 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.466667 | |
| I0330 14:01:58.030427 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58185 (* 0.3 = 0.774556 loss) | |
| I0330 14:01:58.030441 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.73151 (* 0.3 = 0.219453 loss) | |
| I0330 14:01:58.030454 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.288889 | |
| I0330 14:01:58.030467 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0330 14:01:58.030483 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0330 14:01:58.030496 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.25506 (* 0.3 = 0.676517 loss) | |
| I0330 14:01:58.030511 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.648157 (* 0.3 = 0.194447 loss) | |
| I0330 14:01:58.030524 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.422222 | |
| I0330 14:01:58.030536 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 14:01:58.030555 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0330 14:01:58.030570 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87515 (* 1 = 1.87515 loss) | |
| I0330 14:01:58.030583 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.506456 (* 1 = 0.506456 loss) | |
| I0330 14:01:58.030596 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 14:01:58.030608 13762 solver.cpp:245] Train net output #16: total_confidence = 0.171553 | |
| I0330 14:01:58.030624 13762 sgd_solver.cpp:106] Iteration 47500, lr = 0.01 | |
| I0330 14:04:06.964495 13762 solver.cpp:229] Iteration 48000, loss = 2.62056 | |
| I0330 14:04:06.964604 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.363636 | |
| I0330 14:04:06.964624 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 14:04:06.964637 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0330 14:04:06.964653 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80743 (* 0.3 = 0.542229 loss) | |
| I0330 14:04:06.964668 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.498704 (* 0.3 = 0.149611 loss) | |
| I0330 14:04:06.964681 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.613636 | |
| I0330 14:04:06.964694 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 14:04:06.964705 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.863636 | |
| I0330 14:04:06.964720 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.20594 (* 0.3 = 0.361783 loss) | |
| I0330 14:04:06.964735 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.359665 (* 0.3 = 0.107899 loss) | |
| I0330 14:04:06.964746 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.795455 | |
| I0330 14:04:06.964758 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 14:04:06.964771 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0330 14:04:06.964786 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.675141 (* 1 = 0.675141 loss) | |
| I0330 14:04:06.964815 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.193581 (* 1 = 0.193581 loss) | |
| I0330 14:04:06.964828 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:04:06.964840 13762 solver.cpp:245] Train net output #16: total_confidence = 0.444189 | |
| I0330 14:04:06.964853 13762 sgd_solver.cpp:106] Iteration 48000, lr = 0.01 | |
| I0330 14:06:15.649301 13762 solver.cpp:229] Iteration 48500, loss = 2.56446 | |
| I0330 14:06:15.649482 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.511111 | |
| I0330 14:06:15.649503 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 14:06:15.649518 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.844444 | |
| I0330 14:06:15.649534 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.68889 (* 0.3 = 0.506668 loss) | |
| I0330 14:06:15.649549 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.521889 (* 0.3 = 0.156567 loss) | |
| I0330 14:06:15.649564 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.688889 | |
| I0330 14:06:15.649575 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0330 14:06:15.649588 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.933333 | |
| I0330 14:06:15.649603 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.970706 (* 0.3 = 0.291212 loss) | |
| I0330 14:06:15.649618 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.314276 (* 0.3 = 0.0942828 loss) | |
| I0330 14:06:15.649631 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0330 14:06:15.649644 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 14:06:15.649657 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 14:06:15.649672 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.461875 (* 1 = 0.461875 loss) | |
| I0330 14:06:15.649685 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.135743 (* 1 = 0.135743 loss) | |
| I0330 14:06:15.649698 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:06:15.649710 13762 solver.cpp:245] Train net output #16: total_confidence = 0.306345 | |
| I0330 14:06:15.649724 13762 sgd_solver.cpp:106] Iteration 48500, lr = 0.01 | |
| I0330 14:08:24.390120 13762 solver.cpp:229] Iteration 49000, loss = 2.58868 | |
| I0330 14:08:24.390226 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.276596 | |
| I0330 14:08:24.390247 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 14:08:24.390259 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.553191 | |
| I0330 14:08:24.390275 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25743 (* 0.3 = 0.677228 loss) | |
| I0330 14:08:24.390290 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.63354 (* 0.3 = 0.190062 loss) | |
| I0330 14:08:24.390303 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.425532 | |
| I0330 14:08:24.390316 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 14:08:24.390327 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.702128 | |
| I0330 14:08:24.390341 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83912 (* 0.3 = 0.551735 loss) | |
| I0330 14:08:24.390357 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.5288 (* 0.3 = 0.15864 loss) | |
| I0330 14:08:24.390368 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.702128 | |
| I0330 14:08:24.390380 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 14:08:24.390399 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0330 14:08:24.390414 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00743 (* 1 = 1.00743 loss) | |
| I0330 14:08:24.390429 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.294266 (* 1 = 0.294266 loss) | |
| I0330 14:08:24.390440 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 14:08:24.390452 13762 solver.cpp:245] Train net output #16: total_confidence = 0.259034 | |
| I0330 14:08:24.390465 13762 sgd_solver.cpp:106] Iteration 49000, lr = 0.01 | |
| I0330 14:10:33.271776 13762 solver.cpp:229] Iteration 49500, loss = 2.58207 | |
| I0330 14:10:33.271920 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.423077 | |
| I0330 14:10:33.271941 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 14:10:33.271957 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.557692 | |
| I0330 14:10:33.271986 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13776 (* 0.3 = 0.641328 loss) | |
| I0330 14:10:33.272003 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.67905 (* 0.3 = 0.203715 loss) | |
| I0330 14:10:33.272017 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.538462 | |
| I0330 14:10:33.272028 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 14:10:33.272042 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.769231 | |
| I0330 14:10:33.272055 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.51462 (* 0.3 = 0.454387 loss) | |
| I0330 14:10:33.272069 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.471525 (* 0.3 = 0.141457 loss) | |
| I0330 14:10:33.272083 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.711538 | |
| I0330 14:10:33.272094 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 14:10:33.272106 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.903846 | |
| I0330 14:10:33.272121 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.950208 (* 1 = 0.950208 loss) | |
| I0330 14:10:33.272135 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.314231 (* 1 = 0.314231 loss) | |
| I0330 14:10:33.272147 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:10:33.272162 13762 solver.cpp:245] Train net output #16: total_confidence = 0.273655 | |
| I0330 14:10:33.272176 13762 sgd_solver.cpp:106] Iteration 49500, lr = 0.01 | |
| I0330 14:12:42.656961 13762 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_50000.caffemodel | |
| I0330 14:12:43.476646 13762 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm8_bn_iter_50000.solverstate | |
| I0330 14:12:43.643676 13762 solver.cpp:338] Iteration 50000, Testing net (#0) | |
| I0330 14:13:13.596617 13762 solver.cpp:393] Test loss: 2.16944 | |
| I0330 14:13:13.596772 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.526218 | |
| I0330 14:13:13.596791 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.86823 | |
| I0330 14:13:13.596806 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.787769 | |
| I0330 14:13:13.596829 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.61837 (* 0.3 = 0.48551 loss) | |
| I0330 14:13:13.596844 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.452822 (* 0.3 = 0.135846 loss) | |
| I0330 14:13:13.596858 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.689257 | |
| I0330 14:13:13.596870 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.905002 | |
| I0330 14:13:13.596882 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.873959 | |
| I0330 14:13:13.596896 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.14361 (* 0.3 = 0.343083 loss) | |
| I0330 14:13:13.596911 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.341338 (* 0.3 = 0.102401 loss) | |
| I0330 14:13:13.596923 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.787461 | |
| I0330 14:13:13.596935 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.939637 | |
| I0330 14:13:13.596956 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.901143 | |
| I0330 14:13:13.596971 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.860884 (* 1 = 0.860884 loss) | |
| I0330 14:13:13.596984 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.24172 (* 1 = 0.24172 loss) | |
| I0330 14:13:13.596997 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.436 | |
| I0330 14:13:13.597013 13762 solver.cpp:406] Test net output #16: total_confidence = 0.406942 | |
| I0330 14:13:13.748746 13762 solver.cpp:229] Iteration 50000, loss = 2.68447 | |
| I0330 14:13:13.748785 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.534884 | |
| I0330 14:13:13.748802 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0330 14:13:13.748814 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.767442 | |
| I0330 14:13:13.748831 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.70177 (* 0.3 = 0.51053 loss) | |
| I0330 14:13:13.748845 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.546025 (* 0.3 = 0.163807 loss) | |
| I0330 14:13:13.748858 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.604651 | |
| I0330 14:13:13.748872 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 14:13:13.748884 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.837209 | |
| I0330 14:13:13.748898 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.30136 (* 0.3 = 0.390408 loss) | |
| I0330 14:13:13.748913 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.42615 (* 0.3 = 0.127845 loss) | |
| I0330 14:13:13.748925 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.860465 | |
| I0330 14:13:13.748939 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 14:13:13.748950 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.953488 | |
| I0330 14:13:13.748965 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.504099 (* 1 = 0.504099 loss) | |
| I0330 14:13:13.748980 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.141038 (* 1 = 0.141038 loss) | |
| I0330 14:13:13.748991 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:13:13.749003 13762 solver.cpp:245] Train net output #16: total_confidence = 0.386489 | |
| I0330 14:13:13.749025 13762 sgd_solver.cpp:106] Iteration 50000, lr = 0.01 | |
| I0330 14:15:22.596278 13762 solver.cpp:229] Iteration 50500, loss = 2.60937 | |
| I0330 14:15:22.596462 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0330 14:15:22.596484 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 14:15:22.596498 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 14:15:22.596515 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93392 (* 0.3 = 0.580177 loss) | |
| I0330 14:15:22.596530 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.727549 (* 0.3 = 0.218265 loss) | |
| I0330 14:15:22.596544 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.619048 | |
| I0330 14:15:22.596556 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 14:15:22.596568 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.785714 | |
| I0330 14:15:22.596583 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47333 (* 0.3 = 0.441998 loss) | |
| I0330 14:15:22.596597 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.666063 (* 0.3 = 0.199819 loss) | |
| I0330 14:15:22.596611 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.809524 | |
| I0330 14:15:22.596623 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 14:15:22.596635 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.904762 | |
| I0330 14:15:22.596650 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.662178 (* 1 = 0.662178 loss) | |
| I0330 14:15:22.596664 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.379667 (* 1 = 0.379667 loss) | |
| I0330 14:15:22.596678 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 14:15:22.596690 13762 solver.cpp:245] Train net output #16: total_confidence = 0.329149 | |
| I0330 14:15:22.596703 13762 sgd_solver.cpp:106] Iteration 50500, lr = 0.01 | |
| I0330 14:17:31.649807 13762 solver.cpp:229] Iteration 51000, loss = 2.62132 | |
| I0330 14:17:31.649935 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.615385 | |
| I0330 14:17:31.649955 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.886364 | |
| I0330 14:17:31.649968 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.820513 | |
| I0330 14:17:31.649984 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.32684 (* 0.3 = 0.398053 loss) | |
| I0330 14:17:31.650001 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.412956 (* 0.3 = 0.123887 loss) | |
| I0330 14:17:31.650013 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.641026 | |
| I0330 14:17:31.650027 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 14:17:31.650038 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.948718 | |
| I0330 14:17:31.650053 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.970123 (* 0.3 = 0.291037 loss) | |
| I0330 14:17:31.650068 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.30899 (* 0.3 = 0.092697 loss) | |
| I0330 14:17:31.650080 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.794872 | |
| I0330 14:17:31.650092 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 14:17:31.650105 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.974359 | |
| I0330 14:17:31.650120 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.527389 (* 1 = 0.527389 loss) | |
| I0330 14:17:31.650135 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.143386 (* 1 = 0.143386 loss) | |
| I0330 14:17:31.650147 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 14:17:31.650161 13762 solver.cpp:245] Train net output #16: total_confidence = 0.390801 | |
| I0330 14:17:31.650174 13762 sgd_solver.cpp:106] Iteration 51000, lr = 0.01 | |
| I0330 14:19:40.881247 13762 solver.cpp:229] Iteration 51500, loss = 2.65032 | |
| I0330 14:19:40.881392 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0330 14:19:40.881412 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 14:19:40.881427 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 14:19:40.881443 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.99868 (* 0.3 = 0.599604 loss) | |
| I0330 14:19:40.881458 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.623969 (* 0.3 = 0.187191 loss) | |
| I0330 14:19:40.881475 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.583333 | |
| I0330 14:19:40.881489 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 14:19:40.881501 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0330 14:19:40.881515 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.59332 (* 0.3 = 0.477997 loss) | |
| I0330 14:19:40.881531 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.479273 (* 0.3 = 0.143782 loss) | |
| I0330 14:19:40.881542 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.770833 | |
| I0330 14:19:40.881554 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 14:19:40.881567 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.895833 | |
| I0330 14:19:40.881582 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.842215 (* 1 = 0.842215 loss) | |
| I0330 14:19:40.881595 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.289808 (* 1 = 0.289808 loss) | |
| I0330 14:19:40.881608 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:19:40.881620 13762 solver.cpp:245] Train net output #16: total_confidence = 0.310781 | |
| I0330 14:19:40.881633 13762 sgd_solver.cpp:106] Iteration 51500, lr = 0.01 | |
| I0330 14:21:49.914350 13762 solver.cpp:229] Iteration 52000, loss = 2.60997 | |
| I0330 14:21:49.914496 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.37037 | |
| I0330 14:21:49.914517 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 14:21:49.914530 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.722222 | |
| I0330 14:21:49.914547 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86342 (* 0.3 = 0.559026 loss) | |
| I0330 14:21:49.914562 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.574271 (* 0.3 = 0.172281 loss) | |
| I0330 14:21:49.914575 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.462963 | |
| I0330 14:21:49.914588 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 14:21:49.914600 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.814815 | |
| I0330 14:21:49.914614 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.70314 (* 0.3 = 0.510942 loss) | |
| I0330 14:21:49.914629 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524899 (* 0.3 = 0.15747 loss) | |
| I0330 14:21:49.914641 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0330 14:21:49.914654 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 14:21:49.914666 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.925926 | |
| I0330 14:21:49.914680 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.605539 (* 1 = 0.605539 loss) | |
| I0330 14:21:49.914695 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.195408 (* 1 = 0.195408 loss) | |
| I0330 14:21:49.914707 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 14:21:49.914721 13762 solver.cpp:245] Train net output #16: total_confidence = 0.362125 | |
| I0330 14:21:49.914732 13762 sgd_solver.cpp:106] Iteration 52000, lr = 0.01 | |
| I0330 14:23:58.709033 13762 solver.cpp:229] Iteration 52500, loss = 2.66038 | |
| I0330 14:23:58.709238 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.345455 | |
| I0330 14:23:58.709259 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0330 14:23:58.709272 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509091 | |
| I0330 14:23:58.709290 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46915 (* 0.3 = 0.740745 loss) | |
| I0330 14:23:58.709313 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.811247 (* 0.3 = 0.243374 loss) | |
| I0330 14:23:58.709333 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.418182 | |
| I0330 14:23:58.709347 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 14:23:58.709359 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0330 14:23:58.709373 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03457 (* 0.3 = 0.610372 loss) | |
| I0330 14:23:58.709388 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.659993 (* 0.3 = 0.197998 loss) | |
| I0330 14:23:58.709403 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6 | |
| I0330 14:23:58.709425 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 14:23:58.709446 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.781818 | |
| I0330 14:23:58.709476 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.50975 (* 1 = 1.50975 loss) | |
| I0330 14:23:58.709506 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.508649 (* 1 = 0.508649 loss) | |
| I0330 14:23:58.709525 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 14:23:58.709538 13762 solver.cpp:245] Train net output #16: total_confidence = 0.303262 | |
| I0330 14:23:58.709552 13762 sgd_solver.cpp:106] Iteration 52500, lr = 0.01 | |
| I0330 14:26:07.456095 13762 solver.cpp:229] Iteration 53000, loss = 2.56629 | |
| I0330 14:26:07.456215 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.425532 | |
| I0330 14:26:07.456236 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 14:26:07.456249 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.723404 | |
| I0330 14:26:07.456265 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73083 (* 0.3 = 0.519249 loss) | |
| I0330 14:26:07.456279 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.500471 (* 0.3 = 0.150141 loss) | |
| I0330 14:26:07.456292 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.553191 | |
| I0330 14:26:07.456305 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0330 14:26:07.456317 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.893617 | |
| I0330 14:26:07.456331 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.27882 (* 0.3 = 0.383647 loss) | |
| I0330 14:26:07.456346 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.372611 (* 0.3 = 0.111783 loss) | |
| I0330 14:26:07.456359 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.87234 | |
| I0330 14:26:07.456372 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0330 14:26:07.456383 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.978723 | |
| I0330 14:26:07.456398 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.461032 (* 1 = 0.461032 loss) | |
| I0330 14:26:07.456418 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.127012 (* 1 = 0.127012 loss) | |
| I0330 14:26:07.456432 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:26:07.456444 13762 solver.cpp:245] Train net output #16: total_confidence = 0.37253 | |
| I0330 14:26:07.456457 13762 sgd_solver.cpp:106] Iteration 53000, lr = 0.01 | |
| I0330 14:28:16.403609 13762 solver.cpp:229] Iteration 53500, loss = 2.60372 | |
| I0330 14:28:16.403744 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0330 14:28:16.403766 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 14:28:16.403784 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 14:28:16.403806 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.73066 (* 0.3 = 0.819199 loss) | |
| I0330 14:28:16.403822 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.762927 (* 0.3 = 0.228878 loss) | |
| I0330 14:28:16.403836 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377778 | |
| I0330 14:28:16.403847 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 14:28:16.403861 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.688889 | |
| I0330 14:28:16.403874 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.73762 (* 0.3 = 0.821285 loss) | |
| I0330 14:28:16.403889 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.758596 (* 0.3 = 0.227579 loss) | |
| I0330 14:28:16.403901 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.622222 | |
| I0330 14:28:16.403914 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 14:28:16.403926 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0330 14:28:16.403940 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.07045 (* 1 = 2.07045 loss) | |
| I0330 14:28:16.403954 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.5447 (* 1 = 0.5447 loss) | |
| I0330 14:28:16.403966 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:28:16.403978 13762 solver.cpp:245] Train net output #16: total_confidence = 0.410765 | |
| I0330 14:28:16.403992 13762 sgd_solver.cpp:106] Iteration 53500, lr = 0.01 | |
| I0330 14:30:25.269155 13762 solver.cpp:229] Iteration 54000, loss = 2.61384 | |
| I0330 14:30:25.269281 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.354167 | |
| I0330 14:30:25.269302 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 14:30:25.269316 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.541667 | |
| I0330 14:30:25.269338 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.66226 (* 0.3 = 0.798679 loss) | |
| I0330 14:30:25.269353 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.754768 (* 0.3 = 0.22643 loss) | |
| I0330 14:30:25.269366 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.395833 | |
| I0330 14:30:25.269379 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 14:30:25.269397 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.583333 | |
| I0330 14:30:25.269412 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.48986 (* 0.3 = 0.746959 loss) | |
| I0330 14:30:25.269425 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.701101 (* 0.3 = 0.21033 loss) | |
| I0330 14:30:25.269438 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.583333 | |
| I0330 14:30:25.269450 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0330 14:30:25.269462 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0330 14:30:25.269476 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.86862 (* 1 = 1.86862 loss) | |
| I0330 14:30:25.269491 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.528002 (* 1 = 0.528002 loss) | |
| I0330 14:30:25.269503 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:30:25.269515 13762 solver.cpp:245] Train net output #16: total_confidence = 0.202007 | |
| I0330 14:30:25.269528 13762 sgd_solver.cpp:106] Iteration 54000, lr = 0.01 | |
| I0330 14:32:34.350710 13762 solver.cpp:229] Iteration 54500, loss = 2.56661 | |
| I0330 14:32:34.350853 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.45098 | |
| I0330 14:32:34.350874 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 14:32:34.350888 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.686275 | |
| I0330 14:32:34.350903 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.91687 (* 0.3 = 0.57506 loss) | |
| I0330 14:32:34.350926 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.600622 (* 0.3 = 0.180187 loss) | |
| I0330 14:32:34.350939 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.490196 | |
| I0330 14:32:34.350951 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 14:32:34.350965 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.803922 | |
| I0330 14:32:34.350994 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47108 (* 0.3 = 0.441325 loss) | |
| I0330 14:32:34.351011 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.459951 (* 0.3 = 0.137985 loss) | |
| I0330 14:32:34.351023 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0330 14:32:34.351048 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 14:32:34.351061 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.862745 | |
| I0330 14:32:34.351075 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00209 (* 1 = 1.00209 loss) | |
| I0330 14:32:34.351089 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.301959 (* 1 = 0.301959 loss) | |
| I0330 14:32:34.351109 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 14:32:34.351121 13762 solver.cpp:245] Train net output #16: total_confidence = 0.323506 | |
| I0330 14:32:34.351133 13762 sgd_solver.cpp:106] Iteration 54500, lr = 0.01 | |
| I0330 14:34:43.070919 13762 solver.cpp:338] Iteration 55000, Testing net (#0) | |
| I0330 14:35:12.980232 13762 solver.cpp:393] Test loss: 2.28812 | |
| I0330 14:35:12.980278 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.501589 | |
| I0330 14:35:12.980294 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.871094 | |
| I0330 14:35:12.980307 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.760556 | |
| I0330 14:35:12.980324 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.74667 (* 0.3 = 0.524001 loss) | |
| I0330 14:35:12.980337 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.457577 (* 0.3 = 0.137273 loss) | |
| I0330 14:35:12.980350 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.666406 | |
| I0330 14:35:12.980362 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.913684 | |
| I0330 14:35:12.980373 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.862315 | |
| I0330 14:35:12.980387 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.21779 (* 0.3 = 0.365337 loss) | |
| I0330 14:35:12.980402 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.319563 (* 0.3 = 0.095869 loss) | |
| I0330 14:35:12.980414 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.774469 | |
| I0330 14:35:12.980427 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.940092 | |
| I0330 14:35:12.980438 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.888113 | |
| I0330 14:35:12.980468 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.919441 (* 1 = 0.919441 loss) | |
| I0330 14:35:12.980482 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.246205 (* 1 = 0.246205 loss) | |
| I0330 14:35:12.980494 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.445 | |
| I0330 14:35:12.980505 13762 solver.cpp:406] Test net output #16: total_confidence = 0.422299 | |
| I0330 14:35:13.131669 13762 solver.cpp:229] Iteration 55000, loss = 2.56244 | |
| I0330 14:35:13.131805 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.522727 | |
| I0330 14:35:13.131825 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0330 14:35:13.131839 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0330 14:35:13.131853 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.55237 (* 0.3 = 0.46571 loss) | |
| I0330 14:35:13.131868 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.454573 (* 0.3 = 0.136372 loss) | |
| I0330 14:35:13.131881 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.727273 | |
| I0330 14:35:13.131894 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.914773 | |
| I0330 14:35:13.131906 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.863636 | |
| I0330 14:35:13.131921 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.955527 (* 0.3 = 0.286658 loss) | |
| I0330 14:35:13.131934 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.2802 (* 0.3 = 0.0840601 loss) | |
| I0330 14:35:13.131947 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.909091 | |
| I0330 14:35:13.131959 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0330 14:35:13.131971 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0330 14:35:13.131984 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.312216 (* 1 = 0.312216 loss) | |
| I0330 14:35:13.132005 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0807796 (* 1 = 0.0807796 loss) | |
| I0330 14:35:13.132024 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 14:35:13.132035 13762 solver.cpp:245] Train net output #16: total_confidence = 0.649678 | |
| I0330 14:35:13.132047 13762 sgd_solver.cpp:106] Iteration 55000, lr = 0.01 | |
| I0330 14:37:21.879528 13762 solver.cpp:229] Iteration 55500, loss = 2.60395 | |
| I0330 14:37:21.879659 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0330 14:37:21.879679 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 14:37:21.879693 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.444444 | |
| I0330 14:37:21.879709 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.59416 (* 0.3 = 1.07825 loss) | |
| I0330 14:37:21.879724 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.04928 (* 0.3 = 0.314785 loss) | |
| I0330 14:37:21.879736 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0330 14:37:21.879748 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 14:37:21.879761 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.555556 | |
| I0330 14:37:21.879776 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.24886 (* 0.3 = 0.974657 loss) | |
| I0330 14:37:21.879789 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.931103 (* 0.3 = 0.279331 loss) | |
| I0330 14:37:21.879801 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.444444 | |
| I0330 14:37:21.879813 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0330 14:37:21.879825 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.622222 | |
| I0330 14:37:21.879839 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.73657 (* 1 = 2.73657 loss) | |
| I0330 14:37:21.879853 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.735963 (* 1 = 0.735963 loss) | |
| I0330 14:37:21.879865 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 14:37:21.879878 13762 solver.cpp:245] Train net output #16: total_confidence = 0.208678 | |
| I0330 14:37:21.879890 13762 sgd_solver.cpp:106] Iteration 55500, lr = 0.01 | |
| I0330 14:39:30.741343 13762 solver.cpp:229] Iteration 56000, loss = 2.55008 | |
| I0330 14:39:30.741493 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.487805 | |
| I0330 14:39:30.741513 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0330 14:39:30.741528 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.780488 | |
| I0330 14:39:30.741544 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.72122 (* 0.3 = 0.516365 loss) | |
| I0330 14:39:30.741566 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.495452 (* 0.3 = 0.148636 loss) | |
| I0330 14:39:30.741580 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.658537 | |
| I0330 14:39:30.741592 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 14:39:30.741605 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.878049 | |
| I0330 14:39:30.741618 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.08699 (* 0.3 = 0.326097 loss) | |
| I0330 14:39:30.741632 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.344087 (* 0.3 = 0.103226 loss) | |
| I0330 14:39:30.741646 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.853659 | |
| I0330 14:39:30.741657 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 14:39:30.741668 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.951219 | |
| I0330 14:39:30.741683 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.644249 (* 1 = 0.644249 loss) | |
| I0330 14:39:30.741698 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.212141 (* 1 = 0.212141 loss) | |
| I0330 14:39:30.741709 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:39:30.741721 13762 solver.cpp:245] Train net output #16: total_confidence = 0.35187 | |
| I0330 14:39:30.741734 13762 sgd_solver.cpp:106] Iteration 56000, lr = 0.01 | |
| I0330 14:41:39.460904 13762 solver.cpp:229] Iteration 56500, loss = 2.57649 | |
| I0330 14:41:39.460996 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 14:41:39.461015 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 14:41:39.461029 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.622222 | |
| I0330 14:41:39.461045 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14759 (* 0.3 = 0.644278 loss) | |
| I0330 14:41:39.461060 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.581641 (* 0.3 = 0.174492 loss) | |
| I0330 14:41:39.461072 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.6 | |
| I0330 14:41:39.461084 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 14:41:39.461097 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0330 14:41:39.461110 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53682 (* 0.3 = 0.461047 loss) | |
| I0330 14:41:39.461124 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.444039 (* 0.3 = 0.133212 loss) | |
| I0330 14:41:39.461138 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.644444 | |
| I0330 14:41:39.461149 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 14:41:39.461161 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.733333 | |
| I0330 14:41:39.461175 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49469 (* 1 = 1.49469 loss) | |
| I0330 14:41:39.461189 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.407171 (* 1 = 0.407171 loss) | |
| I0330 14:41:39.461201 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 14:41:39.461213 13762 solver.cpp:245] Train net output #16: total_confidence = 0.354535 | |
| I0330 14:41:39.461225 13762 sgd_solver.cpp:106] Iteration 56500, lr = 0.01 | |
| I0330 14:43:48.262647 13762 solver.cpp:229] Iteration 57000, loss = 2.60952 | |
| I0330 14:43:48.262804 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0330 14:43:48.262825 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0330 14:43:48.262838 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489796 | |
| I0330 14:43:48.262855 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76026 (* 0.3 = 0.828078 loss) | |
| I0330 14:43:48.262877 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.833459 (* 0.3 = 0.250038 loss) | |
| I0330 14:43:48.262890 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.346939 | |
| I0330 14:43:48.262903 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0330 14:43:48.262915 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.612245 | |
| I0330 14:43:48.262930 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.66912 (* 0.3 = 0.800736 loss) | |
| I0330 14:43:48.262944 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.816573 (* 0.3 = 0.244972 loss) | |
| I0330 14:43:48.262956 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0330 14:43:48.262969 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0330 14:43:48.262995 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.77551 | |
| I0330 14:43:48.263012 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.72017 (* 1 = 1.72017 loss) | |
| I0330 14:43:48.263026 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.505961 (* 1 = 0.505961 loss) | |
| I0330 14:43:48.263038 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:43:48.263051 13762 solver.cpp:245] Train net output #16: total_confidence = 0.161971 | |
| I0330 14:43:48.263063 13762 sgd_solver.cpp:106] Iteration 57000, lr = 0.01 | |
| I0330 14:45:57.146282 13762 solver.cpp:229] Iteration 57500, loss = 2.59595 | |
| I0330 14:45:57.146452 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.378378 | |
| I0330 14:45:57.146473 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 14:45:57.146486 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.540541 | |
| I0330 14:45:57.146503 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.47844 (* 0.3 = 0.743531 loss) | |
| I0330 14:45:57.146519 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.689838 (* 0.3 = 0.206951 loss) | |
| I0330 14:45:57.146531 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.378378 | |
| I0330 14:45:57.146543 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 14:45:57.146556 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.540541 | |
| I0330 14:45:57.146570 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.46641 (* 0.3 = 0.739922 loss) | |
| I0330 14:45:57.146585 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.688749 (* 0.3 = 0.206625 loss) | |
| I0330 14:45:57.146597 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.594595 | |
| I0330 14:45:57.146610 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 14:45:57.146622 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.675676 | |
| I0330 14:45:57.146636 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.71841 (* 1 = 1.71841 loss) | |
| I0330 14:45:57.146651 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.431569 (* 1 = 0.431569 loss) | |
| I0330 14:45:57.146663 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:45:57.146677 13762 solver.cpp:245] Train net output #16: total_confidence = 0.276165 | |
| I0330 14:45:57.146688 13762 sgd_solver.cpp:106] Iteration 57500, lr = 0.01 | |
| I0330 14:48:05.993935 13762 solver.cpp:229] Iteration 58000, loss = 2.56991 | |
| I0330 14:48:05.994088 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.357143 | |
| I0330 14:48:05.994109 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 14:48:05.994122 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0330 14:48:05.994139 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.0483 (* 0.3 = 0.61449 loss) | |
| I0330 14:48:05.994153 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.688019 (* 0.3 = 0.206406 loss) | |
| I0330 14:48:05.994169 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.47619 | |
| I0330 14:48:05.994182 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0330 14:48:05.994194 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.785714 | |
| I0330 14:48:05.994209 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.56994 (* 0.3 = 0.470983 loss) | |
| I0330 14:48:05.994222 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.578942 (* 0.3 = 0.173683 loss) | |
| I0330 14:48:05.994235 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.761905 | |
| I0330 14:48:05.994248 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 14:48:05.994261 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0330 14:48:05.994276 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.765467 (* 1 = 0.765467 loss) | |
| I0330 14:48:05.994289 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.26683 (* 1 = 0.26683 loss) | |
| I0330 14:48:05.994302 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 14:48:05.994313 13762 solver.cpp:245] Train net output #16: total_confidence = 0.291045 | |
| I0330 14:48:05.994326 13762 sgd_solver.cpp:106] Iteration 58000, lr = 0.01 | |
| I0330 14:50:14.994885 13762 solver.cpp:229] Iteration 58500, loss = 2.48509 | |
| I0330 14:50:14.995030 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.577778 | |
| I0330 14:50:14.995051 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.875 | |
| I0330 14:50:14.995065 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0330 14:50:14.995082 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.48971 (* 0.3 = 0.446914 loss) | |
| I0330 14:50:14.995098 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.433491 (* 0.3 = 0.130047 loss) | |
| I0330 14:50:14.995111 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.688889 | |
| I0330 14:50:14.995123 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.909091 | |
| I0330 14:50:14.995136 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.888889 | |
| I0330 14:50:14.995151 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.04874 (* 0.3 = 0.314621 loss) | |
| I0330 14:50:14.995168 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.306032 (* 0.3 = 0.0918096 loss) | |
| I0330 14:50:14.995182 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.933333 | |
| I0330 14:50:14.995194 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0330 14:50:14.995206 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.955556 | |
| I0330 14:50:14.995221 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.377944 (* 1 = 0.377944 loss) | |
| I0330 14:50:14.995235 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.134387 (* 1 = 0.134387 loss) | |
| I0330 14:50:14.995249 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 14:50:14.995260 13762 solver.cpp:245] Train net output #16: total_confidence = 0.482537 | |
| I0330 14:50:14.995273 13762 sgd_solver.cpp:106] Iteration 58500, lr = 0.01 | |
| I0330 14:52:23.781198 13762 solver.cpp:229] Iteration 59000, loss = 2.60594 | |
| I0330 14:52:23.781365 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.413043 | |
| I0330 14:52:23.781394 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 14:52:23.781407 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.76087 | |
| I0330 14:52:23.781424 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.85194 (* 0.3 = 0.555582 loss) | |
| I0330 14:52:23.781440 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.536018 (* 0.3 = 0.160805 loss) | |
| I0330 14:52:23.781451 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.608696 | |
| I0330 14:52:23.781464 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 14:52:23.781477 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.847826 | |
| I0330 14:52:23.781491 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.26701 (* 0.3 = 0.380103 loss) | |
| I0330 14:52:23.781507 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.360795 (* 0.3 = 0.108239 loss) | |
| I0330 14:52:23.781518 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.804348 | |
| I0330 14:52:23.781530 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 14:52:23.781543 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0330 14:52:23.781558 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.65426 (* 1 = 0.65426 loss) | |
| I0330 14:52:23.781571 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.187003 (* 1 = 0.187003 loss) | |
| I0330 14:52:23.781584 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 14:52:23.781596 13762 solver.cpp:245] Train net output #16: total_confidence = 0.494097 | |
| I0330 14:52:23.781608 13762 sgd_solver.cpp:106] Iteration 59000, lr = 0.01 | |
| I0330 14:54:32.775909 13762 solver.cpp:229] Iteration 59500, loss = 2.55026 | |
| I0330 14:54:32.776033 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.403846 | |
| I0330 14:54:32.776053 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 14:54:32.776067 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.634615 | |
| I0330 14:54:32.776082 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17885 (* 0.3 = 0.653654 loss) | |
| I0330 14:54:32.776098 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.671231 (* 0.3 = 0.201369 loss) | |
| I0330 14:54:32.776109 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.442308 | |
| I0330 14:54:32.776121 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 14:54:32.776134 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 14:54:32.776149 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.69491 (* 0.3 = 0.508474 loss) | |
| I0330 14:54:32.776165 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.546679 (* 0.3 = 0.164004 loss) | |
| I0330 14:54:32.776178 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673077 | |
| I0330 14:54:32.776190 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 14:54:32.776202 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.884615 | |
| I0330 14:54:32.776226 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.04631 (* 1 = 1.04631 loss) | |
| I0330 14:54:32.776265 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.353278 (* 1 = 0.353278 loss) | |
| I0330 14:54:32.776289 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:54:32.776311 13762 solver.cpp:245] Train net output #16: total_confidence = 0.237155 | |
| I0330 14:54:32.776326 13762 sgd_solver.cpp:106] Iteration 59500, lr = 0.01 | |
| I0330 14:56:41.470022 13762 solver.cpp:338] Iteration 60000, Testing net (#0) | |
| I0330 14:57:11.217046 13762 solver.cpp:393] Test loss: 2.11764 | |
| I0330 14:57:11.217092 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.543356 | |
| I0330 14:57:11.217109 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.874684 | |
| I0330 14:57:11.217123 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.800966 | |
| I0330 14:57:11.217139 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.56045 (* 0.3 = 0.468136 loss) | |
| I0330 14:57:11.217152 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.431705 (* 0.3 = 0.129512 loss) | |
| I0330 14:57:11.217167 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.708407 | |
| I0330 14:57:11.217180 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.919865 | |
| I0330 14:57:11.217192 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.881202 | |
| I0330 14:57:11.217206 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.09236 (* 0.3 = 0.327708 loss) | |
| I0330 14:57:11.217221 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.300018 (* 0.3 = 0.0900055 loss) | |
| I0330 14:57:11.217232 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.789026 | |
| I0330 14:57:11.217259 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.944045 | |
| I0330 14:57:11.217272 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.898006 | |
| I0330 14:57:11.217286 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.870671 (* 1 = 0.870671 loss) | |
| I0330 14:57:11.217299 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.231601 (* 1 = 0.231601 loss) | |
| I0330 14:57:11.217311 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.448 | |
| I0330 14:57:11.217322 13762 solver.cpp:406] Test net output #16: total_confidence = 0.448986 | |
| I0330 14:57:11.369004 13762 solver.cpp:229] Iteration 60000, loss = 2.56907 | |
| I0330 14:57:11.369045 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.434783 | |
| I0330 14:57:11.369061 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 14:57:11.369073 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.73913 | |
| I0330 14:57:11.369089 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86462 (* 0.3 = 0.559387 loss) | |
| I0330 14:57:11.369103 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.543763 (* 0.3 = 0.163129 loss) | |
| I0330 14:57:11.369117 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.565217 | |
| I0330 14:57:11.369128 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0330 14:57:11.369140 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0330 14:57:11.369154 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39182 (* 0.3 = 0.417547 loss) | |
| I0330 14:57:11.369168 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.425678 (* 0.3 = 0.127703 loss) | |
| I0330 14:57:11.369180 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.76087 | |
| I0330 14:57:11.369194 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 14:57:11.369220 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.869565 | |
| I0330 14:57:11.369235 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.934932 (* 1 = 0.934932 loss) | |
| I0330 14:57:11.369248 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.264884 (* 1 = 0.264884 loss) | |
| I0330 14:57:11.369261 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 14:57:11.369280 13762 solver.cpp:245] Train net output #16: total_confidence = 0.396006 | |
| I0330 14:57:11.369293 13762 sgd_solver.cpp:106] Iteration 60000, lr = 0.01 | |
| I0330 14:59:20.513617 13762 solver.cpp:229] Iteration 60500, loss = 2.57331 | |
| I0330 14:59:20.513762 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 14:59:20.513783 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 14:59:20.513797 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0330 14:59:20.513813 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11804 (* 0.3 = 0.635413 loss) | |
| I0330 14:59:20.513835 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.601697 (* 0.3 = 0.180509 loss) | |
| I0330 14:59:20.513847 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.555556 | |
| I0330 14:59:20.513860 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 14:59:20.513872 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0330 14:59:20.513887 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76296 (* 0.3 = 0.528888 loss) | |
| I0330 14:59:20.513901 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.493654 (* 0.3 = 0.148096 loss) | |
| I0330 14:59:20.513913 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0330 14:59:20.513928 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 14:59:20.513959 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0330 14:59:20.513998 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.41729 (* 1 = 1.41729 loss) | |
| I0330 14:59:20.514035 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.418243 (* 1 = 0.418243 loss) | |
| I0330 14:59:20.514051 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 14:59:20.514065 13762 solver.cpp:245] Train net output #16: total_confidence = 0.249245 | |
| I0330 14:59:20.514076 13762 sgd_solver.cpp:106] Iteration 60500, lr = 0.01 | |
| I0330 15:01:29.478953 13762 solver.cpp:229] Iteration 61000, loss = 2.51748 | |
| I0330 15:01:29.479138 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.425 | |
| I0330 15:01:29.479161 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 15:01:29.479176 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.675 | |
| I0330 15:01:29.479193 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.77738 (* 0.3 = 0.533213 loss) | |
| I0330 15:01:29.479208 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.516676 (* 0.3 = 0.155003 loss) | |
| I0330 15:01:29.479221 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55 | |
| I0330 15:01:29.479234 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 15:01:29.479248 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 15:01:29.479262 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.485 (* 0.3 = 0.445501 loss) | |
| I0330 15:01:29.479276 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.45122 (* 0.3 = 0.135366 loss) | |
| I0330 15:01:29.479290 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.725 | |
| I0330 15:01:29.479301 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 15:01:29.479313 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.925 | |
| I0330 15:01:29.479327 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.743574 (* 1 = 0.743574 loss) | |
| I0330 15:01:29.479342 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.196843 (* 1 = 0.196843 loss) | |
| I0330 15:01:29.479354 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:01:29.479367 13762 solver.cpp:245] Train net output #16: total_confidence = 0.275079 | |
| I0330 15:01:29.479379 13762 sgd_solver.cpp:106] Iteration 61000, lr = 0.01 | |
| I0330 15:03:38.403888 13762 solver.cpp:229] Iteration 61500, loss = 2.53631 | |
| I0330 15:03:38.404033 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.45 | |
| I0330 15:03:38.404053 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0330 15:03:38.404067 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.675 | |
| I0330 15:03:38.404083 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.86627 (* 0.3 = 0.559882 loss) | |
| I0330 15:03:38.404104 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.479154 (* 0.3 = 0.143746 loss) | |
| I0330 15:03:38.404117 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.475 | |
| I0330 15:03:38.404130 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 15:03:38.404142 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 15:03:38.404156 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.36328 (* 0.3 = 0.408984 loss) | |
| I0330 15:03:38.404175 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.388356 (* 0.3 = 0.116507 loss) | |
| I0330 15:03:38.404187 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.8 | |
| I0330 15:03:38.404199 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 15:03:38.404216 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.925 | |
| I0330 15:03:38.404229 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.774599 (* 1 = 0.774599 loss) | |
| I0330 15:03:38.404243 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.205471 (* 1 = 0.205471 loss) | |
| I0330 15:03:38.404255 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:03:38.404273 13762 solver.cpp:245] Train net output #16: total_confidence = 0.159631 | |
| I0330 15:03:38.404285 13762 sgd_solver.cpp:106] Iteration 61500, lr = 0.01 | |
| I0330 15:05:47.624224 13762 solver.cpp:229] Iteration 62000, loss = 2.55058 | |
| I0330 15:05:47.624343 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0330 15:05:47.624363 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 15:05:47.624377 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72093 | |
| I0330 15:05:47.624393 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.97312 (* 0.3 = 0.591936 loss) | |
| I0330 15:05:47.624408 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.57209 (* 0.3 = 0.171627 loss) | |
| I0330 15:05:47.624421 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0330 15:05:47.624433 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 15:05:47.624445 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.860465 | |
| I0330 15:05:47.624460 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41331 (* 0.3 = 0.423992 loss) | |
| I0330 15:05:47.624475 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.443146 (* 0.3 = 0.132944 loss) | |
| I0330 15:05:47.624486 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.767442 | |
| I0330 15:05:47.624498 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0330 15:05:47.624511 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.906977 | |
| I0330 15:05:47.624524 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.808674 (* 1 = 0.808674 loss) | |
| I0330 15:05:47.624539 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.233812 (* 1 = 0.233812 loss) | |
| I0330 15:05:47.624552 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:05:47.624563 13762 solver.cpp:245] Train net output #16: total_confidence = 0.151657 | |
| I0330 15:05:47.624575 13762 sgd_solver.cpp:106] Iteration 62000, lr = 0.01 | |
| I0330 15:07:56.644434 13762 solver.cpp:229] Iteration 62500, loss = 2.54753 | |
| I0330 15:07:56.644575 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488889 | |
| I0330 15:07:56.644596 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 15:07:56.644609 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.688889 | |
| I0330 15:07:56.644634 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.9454 (* 0.3 = 0.583619 loss) | |
| I0330 15:07:56.644649 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.6055 (* 0.3 = 0.18165 loss) | |
| I0330 15:07:56.644661 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.6 | |
| I0330 15:07:56.644675 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 15:07:56.644687 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0330 15:07:56.644702 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.30446 (* 0.3 = 0.391339 loss) | |
| I0330 15:07:56.644717 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.407592 (* 0.3 = 0.122278 loss) | |
| I0330 15:07:56.644729 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.8 | |
| I0330 15:07:56.644742 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 15:07:56.644753 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.866667 | |
| I0330 15:07:56.644769 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.913181 (* 1 = 0.913181 loss) | |
| I0330 15:07:56.644784 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.239459 (* 1 = 0.239459 loss) | |
| I0330 15:07:56.644796 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:07:56.644809 13762 solver.cpp:245] Train net output #16: total_confidence = 0.468086 | |
| I0330 15:07:56.644829 13762 sgd_solver.cpp:106] Iteration 62500, lr = 0.01 | |
| I0330 15:10:05.673775 13762 solver.cpp:229] Iteration 63000, loss = 2.57019 | |
| I0330 15:10:05.673873 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.38 | |
| I0330 15:10:05.673893 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 15:10:05.673907 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.62 | |
| I0330 15:10:05.673923 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.97737 (* 0.3 = 0.59321 loss) | |
| I0330 15:10:05.673938 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.60743 (* 0.3 = 0.182229 loss) | |
| I0330 15:10:05.673950 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.46 | |
| I0330 15:10:05.673962 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 15:10:05.673974 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0330 15:10:05.673988 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68563 (* 0.3 = 0.505688 loss) | |
| I0330 15:10:05.674003 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.52481 (* 0.3 = 0.157443 loss) | |
| I0330 15:10:05.674015 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.66 | |
| I0330 15:10:05.674027 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0330 15:10:05.674038 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.82 | |
| I0330 15:10:05.674064 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.99493 (* 1 = 0.99493 loss) | |
| I0330 15:10:05.674080 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.31214 (* 1 = 0.31214 loss) | |
| I0330 15:10:05.674099 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:10:05.674123 13762 solver.cpp:245] Train net output #16: total_confidence = 0.368835 | |
| I0330 15:10:05.674135 13762 sgd_solver.cpp:106] Iteration 63000, lr = 0.01 | |
| I0330 15:12:14.567829 13762 solver.cpp:229] Iteration 63500, loss = 2.61112 | |
| I0330 15:12:14.567935 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.522727 | |
| I0330 15:12:14.567955 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 15:12:14.567968 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0330 15:12:14.567984 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.70036 (* 0.3 = 0.510108 loss) | |
| I0330 15:12:14.567999 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.547845 (* 0.3 = 0.164354 loss) | |
| I0330 15:12:14.568012 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.636364 | |
| I0330 15:12:14.568024 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0330 15:12:14.568037 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.818182 | |
| I0330 15:12:14.568050 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39039 (* 0.3 = 0.417118 loss) | |
| I0330 15:12:14.568065 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.4001 (* 0.3 = 0.12003 loss) | |
| I0330 15:12:14.568078 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.954545 | |
| I0330 15:12:14.568090 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0330 15:12:14.568102 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0330 15:12:14.568116 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.282072 (* 1 = 0.282072 loss) | |
| I0330 15:12:14.568142 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0829594 (* 1 = 0.0829594 loss) | |
| I0330 15:12:14.568156 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 15:12:14.568171 13762 solver.cpp:245] Train net output #16: total_confidence = 0.463888 | |
| I0330 15:12:14.568192 13762 sgd_solver.cpp:106] Iteration 63500, lr = 0.01 | |
| I0330 15:14:23.586308 13762 solver.cpp:229] Iteration 64000, loss = 2.62714 | |
| I0330 15:14:23.586454 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.384615 | |
| I0330 15:14:23.586475 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 15:14:23.586488 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.717949 | |
| I0330 15:14:23.586504 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.87066 (* 0.3 = 0.561197 loss) | |
| I0330 15:14:23.586526 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.528828 (* 0.3 = 0.158648 loss) | |
| I0330 15:14:23.586539 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.615385 | |
| I0330 15:14:23.586552 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 15:14:23.586565 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.769231 | |
| I0330 15:14:23.586578 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53879 (* 0.3 = 0.461638 loss) | |
| I0330 15:14:23.586593 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.412112 (* 0.3 = 0.123634 loss) | |
| I0330 15:14:23.586606 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.769231 | |
| I0330 15:14:23.586618 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 15:14:23.586630 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0330 15:14:23.586644 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.994803 (* 1 = 0.994803 loss) | |
| I0330 15:14:23.586658 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.259771 (* 1 = 0.259771 loss) | |
| I0330 15:14:23.586678 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:14:23.586697 13762 solver.cpp:245] Train net output #16: total_confidence = 0.271394 | |
| I0330 15:14:23.586710 13762 sgd_solver.cpp:106] Iteration 64000, lr = 0.01 | |
| I0330 15:16:33.139673 13762 solver.cpp:229] Iteration 64500, loss = 2.54935 | |
| I0330 15:16:33.139787 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.521739 | |
| I0330 15:16:33.139807 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0330 15:16:33.139821 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.695652 | |
| I0330 15:16:33.139837 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.69846 (* 0.3 = 0.509537 loss) | |
| I0330 15:16:33.139852 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.482552 (* 0.3 = 0.144766 loss) | |
| I0330 15:16:33.139864 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.565217 | |
| I0330 15:16:33.139876 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 15:16:33.139889 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.73913 | |
| I0330 15:16:33.139902 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47879 (* 0.3 = 0.443637 loss) | |
| I0330 15:16:33.139916 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.439391 (* 0.3 = 0.131817 loss) | |
| I0330 15:16:33.139928 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.804348 | |
| I0330 15:16:33.139941 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 15:16:33.139953 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.913043 | |
| I0330 15:16:33.139967 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.674739 (* 1 = 0.674739 loss) | |
| I0330 15:16:33.139981 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.199076 (* 1 = 0.199076 loss) | |
| I0330 15:16:33.139993 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:16:33.140012 13762 solver.cpp:245] Train net output #16: total_confidence = 0.443948 | |
| I0330 15:16:33.140049 13762 sgd_solver.cpp:106] Iteration 64500, lr = 0.01 | |
| I0330 15:18:41.848052 13762 solver.cpp:338] Iteration 65000, Testing net (#0) | |
| I0330 15:19:11.751724 13762 solver.cpp:393] Test loss: 2.43842 | |
| I0330 15:19:11.751785 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.401423 | |
| I0330 15:19:11.751801 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.852593 | |
| I0330 15:19:11.751816 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.702572 | |
| I0330 15:19:11.751832 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.11042 (* 0.3 = 0.633127 loss) | |
| I0330 15:19:11.751845 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.527308 (* 0.3 = 0.158192 loss) | |
| I0330 15:19:11.751858 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.67147 | |
| I0330 15:19:11.751870 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.915502 | |
| I0330 15:19:11.751881 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.86219 | |
| I0330 15:19:11.751895 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.21216 (* 0.3 = 0.363649 loss) | |
| I0330 15:19:11.751909 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.315816 (* 0.3 = 0.0947447 loss) | |
| I0330 15:19:11.751921 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.770604 | |
| I0330 15:19:11.751934 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.942501 | |
| I0330 15:19:11.751945 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.884721 | |
| I0330 15:19:11.751960 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.946508 (* 1 = 0.946508 loss) | |
| I0330 15:19:11.751974 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.242192 (* 1 = 0.242192 loss) | |
| I0330 15:19:11.751986 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.446 | |
| I0330 15:19:11.751997 13762 solver.cpp:406] Test net output #16: total_confidence = 0.396742 | |
| I0330 15:19:11.903815 13762 solver.cpp:229] Iteration 65000, loss = 2.62364 | |
| I0330 15:19:11.903921 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.452381 | |
| I0330 15:19:11.903940 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0330 15:19:11.903954 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0330 15:19:11.903970 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.87177 (* 0.3 = 0.56153 loss) | |
| I0330 15:19:11.903985 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.526068 (* 0.3 = 0.157821 loss) | |
| I0330 15:19:11.903996 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.642857 | |
| I0330 15:19:11.904009 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0330 15:19:11.904021 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.857143 | |
| I0330 15:19:11.904036 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.25849 (* 0.3 = 0.377547 loss) | |
| I0330 15:19:11.904049 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.361711 (* 0.3 = 0.108513 loss) | |
| I0330 15:19:11.904062 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.857143 | |
| I0330 15:19:11.904075 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 15:19:11.904088 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 15:19:11.904101 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.384505 (* 1 = 0.384505 loss) | |
| I0330 15:19:11.904115 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.114406 (* 1 = 0.114406 loss) | |
| I0330 15:19:11.904141 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:19:11.904155 13762 solver.cpp:245] Train net output #16: total_confidence = 0.432208 | |
| I0330 15:19:11.904172 13762 sgd_solver.cpp:106] Iteration 65000, lr = 0.01 | |
| I0330 15:21:20.925081 13762 solver.cpp:229] Iteration 65500, loss = 2.51245 | |
| I0330 15:21:20.925240 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.375 | |
| I0330 15:21:20.925261 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 15:21:20.925283 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0330 15:21:20.925299 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.59964 (* 0.3 = 0.779893 loss) | |
| I0330 15:21:20.925315 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.756067 (* 0.3 = 0.22682 loss) | |
| I0330 15:21:20.925328 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 15:21:20.925340 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 15:21:20.925353 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6875 | |
| I0330 15:21:20.925366 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61701 (* 0.3 = 0.485102 loss) | |
| I0330 15:21:20.925380 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.470501 (* 0.3 = 0.14115 loss) | |
| I0330 15:21:20.925393 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 15:21:20.925405 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 15:21:20.925426 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0330 15:21:20.925460 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.915239 (* 1 = 0.915239 loss) | |
| I0330 15:21:20.925488 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.273492 (* 1 = 0.273492 loss) | |
| I0330 15:21:20.925521 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:21:20.925539 13762 solver.cpp:245] Train net output #16: total_confidence = 0.209644 | |
| I0330 15:21:20.925552 13762 sgd_solver.cpp:106] Iteration 65500, lr = 0.01 | |
| I0330 15:23:29.829603 13762 solver.cpp:229] Iteration 66000, loss = 2.52452 | |
| I0330 15:23:29.829782 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0330 15:23:29.829803 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0330 15:23:29.829816 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5625 | |
| I0330 15:23:29.829833 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34774 (* 0.3 = 0.704321 loss) | |
| I0330 15:23:29.829849 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.745596 (* 0.3 = 0.223679 loss) | |
| I0330 15:23:29.829860 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.458333 | |
| I0330 15:23:29.829874 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0330 15:23:29.829885 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0330 15:23:29.829900 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9269 (* 0.3 = 0.578071 loss) | |
| I0330 15:23:29.829915 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.618631 (* 0.3 = 0.185589 loss) | |
| I0330 15:23:29.829927 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 15:23:29.829939 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 15:23:29.829952 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0330 15:23:29.829965 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.876293 (* 1 = 0.876293 loss) | |
| I0330 15:23:29.829980 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.262174 (* 1 = 0.262174 loss) | |
| I0330 15:23:29.829993 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 15:23:29.830004 13762 solver.cpp:245] Train net output #16: total_confidence = 0.207895 | |
| I0330 15:23:29.830018 13762 sgd_solver.cpp:106] Iteration 66000, lr = 0.01 | |
| I0330 15:25:38.973006 13762 solver.cpp:229] Iteration 66500, loss = 2.57927 | |
| I0330 15:25:38.973145 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0330 15:25:38.973166 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0330 15:25:38.973179 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.622222 | |
| I0330 15:25:38.973196 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15901 (* 0.3 = 0.647704 loss) | |
| I0330 15:25:38.973211 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.625915 (* 0.3 = 0.187775 loss) | |
| I0330 15:25:38.973224 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0330 15:25:38.973237 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 15:25:38.973249 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.711111 | |
| I0330 15:25:38.973263 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.8555 (* 0.3 = 0.556651 loss) | |
| I0330 15:25:38.973278 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.527222 (* 0.3 = 0.158167 loss) | |
| I0330 15:25:38.973289 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0330 15:25:38.973301 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0330 15:25:38.973314 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0330 15:25:38.973333 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.02681 (* 1 = 1.02681 loss) | |
| I0330 15:25:38.973348 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.298482 (* 1 = 0.298482 loss) | |
| I0330 15:25:38.973361 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 15:25:38.973373 13762 solver.cpp:245] Train net output #16: total_confidence = 0.208809 | |
| I0330 15:25:38.973386 13762 sgd_solver.cpp:106] Iteration 66500, lr = 0.01 | |
| I0330 15:27:48.015328 13762 solver.cpp:229] Iteration 67000, loss = 2.50867 | |
| I0330 15:27:48.015442 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.212121 | |
| I0330 15:27:48.015462 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0330 15:27:48.015475 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.484848 | |
| I0330 15:27:48.015491 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82467 (* 0.3 = 0.8474 loss) | |
| I0330 15:27:48.015506 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.825477 (* 0.3 = 0.247643 loss) | |
| I0330 15:27:48.015518 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.30303 | |
| I0330 15:27:48.015532 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 15:27:48.015544 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.575758 | |
| I0330 15:27:48.015558 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.44931 (* 0.3 = 0.734794 loss) | |
| I0330 15:27:48.015573 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.72168 (* 0.3 = 0.216504 loss) | |
| I0330 15:27:48.015585 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.484848 | |
| I0330 15:27:48.015597 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0330 15:27:48.015610 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0330 15:27:48.015625 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.30286 (* 1 = 2.30286 loss) | |
| I0330 15:27:48.015638 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.577164 (* 1 = 0.577164 loss) | |
| I0330 15:27:48.015657 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 15:27:48.015677 13762 solver.cpp:245] Train net output #16: total_confidence = 0.134322 | |
| I0330 15:27:48.015691 13762 sgd_solver.cpp:106] Iteration 67000, lr = 0.01 | |
| I0330 15:29:57.023775 13762 solver.cpp:229] Iteration 67500, loss = 2.50025 | |
| I0330 15:29:57.023901 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.292683 | |
| I0330 15:29:57.023921 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 15:29:57.023934 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.731707 | |
| I0330 15:29:57.023952 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02642 (* 0.3 = 0.607925 loss) | |
| I0330 15:29:57.023967 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.645175 (* 0.3 = 0.193553 loss) | |
| I0330 15:29:57.023982 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.390244 | |
| I0330 15:29:57.023994 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0330 15:29:57.024006 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.609756 | |
| I0330 15:29:57.024021 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.07773 (* 0.3 = 0.623319 loss) | |
| I0330 15:29:57.024035 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.638082 (* 0.3 = 0.191425 loss) | |
| I0330 15:29:57.024049 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.634146 | |
| I0330 15:29:57.024060 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0330 15:29:57.024072 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.878049 | |
| I0330 15:29:57.024093 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.16537 (* 1 = 1.16537 loss) | |
| I0330 15:29:57.024107 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.292594 (* 1 = 0.292594 loss) | |
| I0330 15:29:57.024121 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:29:57.024132 13762 solver.cpp:245] Train net output #16: total_confidence = 0.166206 | |
| I0330 15:29:57.024144 13762 sgd_solver.cpp:106] Iteration 67500, lr = 0.01 | |
| I0330 15:32:05.803540 13762 solver.cpp:229] Iteration 68000, loss = 2.4795 | |
| I0330 15:32:05.803702 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.355556 | |
| I0330 15:32:05.803724 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0330 15:32:05.803737 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 15:32:05.803755 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02092 (* 0.3 = 0.606277 loss) | |
| I0330 15:32:05.803769 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.607735 (* 0.3 = 0.18232 loss) | |
| I0330 15:32:05.803781 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.644444 | |
| I0330 15:32:05.803794 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0330 15:32:05.803807 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0330 15:32:05.803820 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53806 (* 0.3 = 0.461419 loss) | |
| I0330 15:32:05.803835 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.470043 (* 0.3 = 0.141013 loss) | |
| I0330 15:32:05.803848 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0330 15:32:05.803860 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 15:32:05.803872 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0330 15:32:05.803886 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.861159 (* 1 = 0.861159 loss) | |
| I0330 15:32:05.803901 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27108 (* 1 = 0.27108 loss) | |
| I0330 15:32:05.803913 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:32:05.803926 13762 solver.cpp:245] Train net output #16: total_confidence = 0.304292 | |
| I0330 15:32:05.803938 13762 sgd_solver.cpp:106] Iteration 68000, lr = 0.01 | |
| I0330 15:34:14.844493 13762 solver.cpp:229] Iteration 68500, loss = 2.50419 | |
| I0330 15:34:14.844645 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.377778 | |
| I0330 15:34:14.844666 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 15:34:14.844678 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0330 15:34:14.844694 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10951 (* 0.3 = 0.632854 loss) | |
| I0330 15:34:14.844709 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.571334 (* 0.3 = 0.1714 loss) | |
| I0330 15:34:14.844722 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0330 15:34:14.844735 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 15:34:14.844753 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.733333 | |
| I0330 15:34:14.844768 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.52445 (* 0.3 = 0.457334 loss) | |
| I0330 15:34:14.844781 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.425241 (* 0.3 = 0.127572 loss) | |
| I0330 15:34:14.844794 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.644444 | |
| I0330 15:34:14.844806 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0330 15:34:14.844818 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0330 15:34:14.844841 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.2208 (* 1 = 1.2208 loss) | |
| I0330 15:34:14.844856 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.342901 (* 1 = 0.342901 loss) | |
| I0330 15:34:14.844867 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 15:34:14.844879 13762 solver.cpp:245] Train net output #16: total_confidence = 0.257251 | |
| I0330 15:34:14.844897 13762 sgd_solver.cpp:106] Iteration 68500, lr = 0.01 | |
| I0330 15:36:23.817664 13762 solver.cpp:229] Iteration 69000, loss = 2.49988 | |
| I0330 15:36:23.817787 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326923 | |
| I0330 15:36:23.817807 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0330 15:36:23.817821 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.576923 | |
| I0330 15:36:23.817836 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48616 (* 0.3 = 0.745848 loss) | |
| I0330 15:36:23.817852 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.768865 (* 0.3 = 0.23066 loss) | |
| I0330 15:36:23.817863 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.403846 | |
| I0330 15:36:23.817876 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0330 15:36:23.817889 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.711538 | |
| I0330 15:36:23.817903 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9785 (* 0.3 = 0.593549 loss) | |
| I0330 15:36:23.817919 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.618909 (* 0.3 = 0.185673 loss) | |
| I0330 15:36:23.817930 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.576923 | |
| I0330 15:36:23.817944 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 15:36:23.817955 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826923 | |
| I0330 15:36:23.817977 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29632 (* 1 = 1.29632 loss) | |
| I0330 15:36:23.817991 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.396224 (* 1 = 0.396224 loss) | |
| I0330 15:36:23.818004 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 15:36:23.818017 13762 solver.cpp:245] Train net output #16: total_confidence = 0.200314 | |
| I0330 15:36:23.818037 13762 sgd_solver.cpp:106] Iteration 69000, lr = 0.01 | |
| I0330 15:38:32.569720 13762 solver.cpp:229] Iteration 69500, loss = 2.52184 | |
| I0330 15:38:32.569874 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404255 | |
| I0330 15:38:32.569895 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 15:38:32.569906 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.638298 | |
| I0330 15:38:32.569923 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02567 (* 0.3 = 0.607701 loss) | |
| I0330 15:38:32.569938 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.57613 (* 0.3 = 0.172839 loss) | |
| I0330 15:38:32.569962 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0330 15:38:32.569974 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0330 15:38:32.569986 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.787234 | |
| I0330 15:38:32.570000 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.51386 (* 0.3 = 0.454157 loss) | |
| I0330 15:38:32.570015 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.425447 (* 0.3 = 0.127634 loss) | |
| I0330 15:38:32.570027 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.702128 | |
| I0330 15:38:32.570039 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 15:38:32.570051 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.893617 | |
| I0330 15:38:32.570065 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.831966 (* 1 = 0.831966 loss) | |
| I0330 15:38:32.570080 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.226314 (* 1 = 0.226314 loss) | |
| I0330 15:38:32.570092 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:38:32.570104 13762 solver.cpp:245] Train net output #16: total_confidence = 0.316404 | |
| I0330 15:38:32.570116 13762 sgd_solver.cpp:106] Iteration 69500, lr = 0.01 | |
| I0330 15:40:41.340241 13762 solver.cpp:338] Iteration 70000, Testing net (#0) | |
| I0330 15:41:11.065060 13762 solver.cpp:393] Test loss: 2.13086 | |
| I0330 15:41:11.065106 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.553779 | |
| I0330 15:41:11.065124 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.879821 | |
| I0330 15:41:11.065137 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.805702 | |
| I0330 15:41:11.065153 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.54511 (* 0.3 = 0.463534 loss) | |
| I0330 15:41:11.065171 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.418701 (* 0.3 = 0.12561 loss) | |
| I0330 15:41:11.065184 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.695357 | |
| I0330 15:41:11.065196 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.918092 | |
| I0330 15:41:11.065208 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.879046 | |
| I0330 15:41:11.065222 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.13056 (* 0.3 = 0.339168 loss) | |
| I0330 15:41:11.065237 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.304965 (* 0.3 = 0.0914896 loss) | |
| I0330 15:41:11.065249 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.782728 | |
| I0330 15:41:11.065261 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.943546 | |
| I0330 15:41:11.065274 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.898489 | |
| I0330 15:41:11.065291 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.876899 (* 1 = 0.876899 loss) | |
| I0330 15:41:11.065305 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.234155 (* 1 = 0.234155 loss) | |
| I0330 15:41:11.065317 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.469 | |
| I0330 15:41:11.065328 13762 solver.cpp:406] Test net output #16: total_confidence = 0.423304 | |
| I0330 15:41:11.216548 13762 solver.cpp:229] Iteration 70000, loss = 2.58977 | |
| I0330 15:41:11.216601 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.477273 | |
| I0330 15:41:11.216619 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0330 15:41:11.216632 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.772727 | |
| I0330 15:41:11.216648 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.64219 (* 0.3 = 0.492658 loss) | |
| I0330 15:41:11.216663 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.522382 (* 0.3 = 0.156715 loss) | |
| I0330 15:41:11.216686 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.681818 | |
| I0330 15:41:11.216722 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 15:41:11.216744 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0330 15:41:11.216778 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.11384 (* 0.3 = 0.334152 loss) | |
| I0330 15:41:11.216801 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.36523 (* 0.3 = 0.109569 loss) | |
| I0330 15:41:11.216821 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.818182 | |
| I0330 15:41:11.216841 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 15:41:11.216861 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.886364 | |
| I0330 15:41:11.216883 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.676044 (* 1 = 0.676044 loss) | |
| I0330 15:41:11.216899 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.200293 (* 1 = 0.200293 loss) | |
| I0330 15:41:11.216912 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 15:41:11.216924 13762 solver.cpp:245] Train net output #16: total_confidence = 0.5633 | |
| I0330 15:41:11.216936 13762 sgd_solver.cpp:106] Iteration 70000, lr = 0.01 | |
| I0330 15:43:20.177585 13762 solver.cpp:229] Iteration 70500, loss = 2.53093 | |
| I0330 15:43:20.177743 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.387755 | |
| I0330 15:43:20.177763 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 15:43:20.177783 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.612245 | |
| I0330 15:43:20.177799 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.16788 (* 0.3 = 0.650365 loss) | |
| I0330 15:43:20.177814 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.645578 (* 0.3 = 0.193673 loss) | |
| I0330 15:43:20.177827 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0330 15:43:20.177839 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 15:43:20.177852 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0330 15:43:20.177866 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.62971 (* 0.3 = 0.488912 loss) | |
| I0330 15:43:20.177881 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.469152 (* 0.3 = 0.140746 loss) | |
| I0330 15:43:20.177896 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0330 15:43:20.177918 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0330 15:43:20.177958 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.918367 | |
| I0330 15:43:20.177997 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.975442 (* 1 = 0.975442 loss) | |
| I0330 15:43:20.178015 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.281481 (* 1 = 0.281481 loss) | |
| I0330 15:43:20.178027 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:43:20.178040 13762 solver.cpp:245] Train net output #16: total_confidence = 0.244296 | |
| I0330 15:43:20.178051 13762 sgd_solver.cpp:106] Iteration 70500, lr = 0.01 | |
| I0330 15:45:29.268473 13762 solver.cpp:229] Iteration 71000, loss = 2.48933 | |
| I0330 15:45:29.268623 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0330 15:45:29.268645 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 15:45:29.268657 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.568627 | |
| I0330 15:45:29.268674 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4097 (* 0.3 = 0.722909 loss) | |
| I0330 15:45:29.268689 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.723243 (* 0.3 = 0.216973 loss) | |
| I0330 15:45:29.268702 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.411765 | |
| I0330 15:45:29.268717 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 15:45:29.268728 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.784314 | |
| I0330 15:45:29.268743 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.81721 (* 0.3 = 0.545163 loss) | |
| I0330 15:45:29.268757 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.556926 (* 0.3 = 0.167078 loss) | |
| I0330 15:45:29.268769 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.803922 | |
| I0330 15:45:29.268781 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 15:45:29.268795 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.960784 | |
| I0330 15:45:29.268808 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.761431 (* 1 = 0.761431 loss) | |
| I0330 15:45:29.268822 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.226297 (* 1 = 0.226297 loss) | |
| I0330 15:45:29.268834 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:45:29.268846 13762 solver.cpp:245] Train net output #16: total_confidence = 0.28673 | |
| I0330 15:45:29.268858 13762 sgd_solver.cpp:106] Iteration 71000, lr = 0.01 | |
| I0330 15:47:38.319548 13762 solver.cpp:229] Iteration 71500, loss = 2.49718 | |
| I0330 15:47:38.319681 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.363636 | |
| I0330 15:47:38.319701 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 15:47:38.319716 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.590909 | |
| I0330 15:47:38.319732 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.38874 (* 0.3 = 0.716622 loss) | |
| I0330 15:47:38.319747 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.656097 (* 0.3 = 0.196829 loss) | |
| I0330 15:47:38.319761 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.295455 | |
| I0330 15:47:38.319772 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0330 15:47:38.319785 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.613636 | |
| I0330 15:47:38.319799 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.3534 (* 0.3 = 0.706019 loss) | |
| I0330 15:47:38.319813 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.638707 (* 0.3 = 0.191612 loss) | |
| I0330 15:47:38.319825 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0330 15:47:38.319838 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0330 15:47:38.319850 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.659091 | |
| I0330 15:47:38.319864 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.06941 (* 1 = 2.06941 loss) | |
| I0330 15:47:38.319880 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.545323 (* 1 = 0.545323 loss) | |
| I0330 15:47:38.319891 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:47:38.319903 13762 solver.cpp:245] Train net output #16: total_confidence = 0.238314 | |
| I0330 15:47:38.319916 13762 sgd_solver.cpp:106] Iteration 71500, lr = 0.01 | |
| I0330 15:49:47.339100 13762 solver.cpp:229] Iteration 72000, loss = 2.49159 | |
| I0330 15:49:47.339242 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.590909 | |
| I0330 15:49:47.339262 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0330 15:49:47.339275 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.772727 | |
| I0330 15:49:47.339293 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.64098 (* 0.3 = 0.492294 loss) | |
| I0330 15:49:47.339308 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.506462 (* 0.3 = 0.151939 loss) | |
| I0330 15:49:47.339320 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.636364 | |
| I0330 15:49:47.339332 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0330 15:49:47.339351 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.863636 | |
| I0330 15:49:47.339365 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41181 (* 0.3 = 0.423542 loss) | |
| I0330 15:49:47.339380 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.413996 (* 0.3 = 0.124199 loss) | |
| I0330 15:49:47.339392 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.840909 | |
| I0330 15:49:47.339409 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0330 15:49:47.339421 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.931818 | |
| I0330 15:49:47.339437 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.975231 (* 1 = 0.975231 loss) | |
| I0330 15:49:47.339450 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.247849 (* 1 = 0.247849 loss) | |
| I0330 15:49:47.339468 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 15:49:47.339479 13762 solver.cpp:245] Train net output #16: total_confidence = 0.572242 | |
| I0330 15:49:47.339491 13762 sgd_solver.cpp:106] Iteration 72000, lr = 0.01 | |
| I0330 15:51:56.477385 13762 solver.cpp:229] Iteration 72500, loss = 2.53068 | |
| I0330 15:51:56.477506 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.232143 | |
| I0330 15:51:56.477525 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0330 15:51:56.477540 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.535714 | |
| I0330 15:51:56.477555 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.43266 (* 0.3 = 0.729799 loss) | |
| I0330 15:51:56.477571 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.790396 (* 0.3 = 0.237119 loss) | |
| I0330 15:51:56.477582 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.446429 | |
| I0330 15:51:56.477596 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0330 15:51:56.477608 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.678571 | |
| I0330 15:51:56.477622 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04464 (* 0.3 = 0.613391 loss) | |
| I0330 15:51:56.477637 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.665209 (* 0.3 = 0.199563 loss) | |
| I0330 15:51:56.477648 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0330 15:51:56.477661 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0330 15:51:56.477672 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0330 15:51:56.477686 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.25891 (* 1 = 1.25891 loss) | |
| I0330 15:51:56.477700 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.414678 (* 1 = 0.414678 loss) | |
| I0330 15:51:56.477713 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:51:56.477725 13762 solver.cpp:245] Train net output #16: total_confidence = 0.268782 | |
| I0330 15:51:56.477737 13762 sgd_solver.cpp:106] Iteration 72500, lr = 0.01 | |
| I0330 15:54:05.504585 13762 solver.cpp:229] Iteration 73000, loss = 2.47107 | |
| I0330 15:54:05.504712 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.363636 | |
| I0330 15:54:05.504732 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0330 15:54:05.504748 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.545455 | |
| I0330 15:54:05.504765 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48153 (* 0.3 = 0.744458 loss) | |
| I0330 15:54:05.504781 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.707054 (* 0.3 = 0.212116 loss) | |
| I0330 15:54:05.504801 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.590909 | |
| I0330 15:54:05.504813 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 15:54:05.504827 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.818182 | |
| I0330 15:54:05.504840 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61835 (* 0.3 = 0.485506 loss) | |
| I0330 15:54:05.504854 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.517037 (* 0.3 = 0.155111 loss) | |
| I0330 15:54:05.504873 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0330 15:54:05.504906 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0330 15:54:05.504931 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795455 | |
| I0330 15:54:05.504968 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.41769 (* 1 = 1.41769 loss) | |
| I0330 15:54:05.504984 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.422206 (* 1 = 0.422206 loss) | |
| I0330 15:54:05.504997 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0330 15:54:05.505009 13762 solver.cpp:245] Train net output #16: total_confidence = 0.309258 | |
| I0330 15:54:05.505022 13762 sgd_solver.cpp:106] Iteration 73000, lr = 0.01 | |
| I0330 15:56:14.493278 13762 solver.cpp:229] Iteration 73500, loss = 2.50859 | |
| I0330 15:56:14.493391 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.367347 | |
| I0330 15:56:14.493410 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 15:56:14.493428 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.693878 | |
| I0330 15:56:14.493445 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78489 (* 0.3 = 0.535468 loss) | |
| I0330 15:56:14.493459 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.547584 (* 0.3 = 0.164275 loss) | |
| I0330 15:56:14.493473 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.673469 | |
| I0330 15:56:14.493485 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0330 15:56:14.493497 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.857143 | |
| I0330 15:56:14.493510 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.14486 (* 0.3 = 0.343458 loss) | |
| I0330 15:56:14.493525 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.355691 (* 0.3 = 0.106707 loss) | |
| I0330 15:56:14.493538 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.836735 | |
| I0330 15:56:14.493551 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 15:56:14.493562 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 15:56:14.493577 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.448295 (* 1 = 0.448295 loss) | |
| I0330 15:56:14.493590 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.140582 (* 1 = 0.140582 loss) | |
| I0330 15:56:14.493602 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 15:56:14.493614 13762 solver.cpp:245] Train net output #16: total_confidence = 0.30956 | |
| I0330 15:56:14.493626 13762 sgd_solver.cpp:106] Iteration 73500, lr = 0.01 | |
| I0330 15:58:23.474100 13762 solver.cpp:229] Iteration 74000, loss = 2.48726 | |
| I0330 15:58:23.474238 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.191489 | |
| I0330 15:58:23.474259 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0330 15:58:23.474272 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.446809 | |
| I0330 15:58:23.474288 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82471 (* 0.3 = 0.847412 loss) | |
| I0330 15:58:23.474309 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.886848 (* 0.3 = 0.266054 loss) | |
| I0330 15:58:23.474323 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.234043 | |
| I0330 15:58:23.474334 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0330 15:58:23.474347 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.489362 | |
| I0330 15:58:23.474361 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.85131 (* 0.3 = 0.855393 loss) | |
| I0330 15:58:23.474375 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.891856 (* 0.3 = 0.267557 loss) | |
| I0330 15:58:23.474393 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.297872 | |
| I0330 15:58:23.474414 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0330 15:58:23.474426 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.595745 | |
| I0330 15:58:23.474441 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.62779 (* 1 = 2.62779 loss) | |
| I0330 15:58:23.474454 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.766765 (* 1 = 0.766765 loss) | |
| I0330 15:58:23.474467 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 15:58:23.474483 13762 solver.cpp:245] Train net output #16: total_confidence = 0.142444 | |
| I0330 15:58:23.474494 13762 sgd_solver.cpp:106] Iteration 74000, lr = 0.01 | |
| I0330 16:00:32.532142 13762 solver.cpp:229] Iteration 74500, loss = 2.47226 | |
| I0330 16:00:32.532238 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.306122 | |
| I0330 16:00:32.532258 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0330 16:00:32.532271 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489796 | |
| I0330 16:00:32.532287 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.71428 (* 0.3 = 0.814285 loss) | |
| I0330 16:00:32.532302 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.790507 (* 0.3 = 0.237152 loss) | |
| I0330 16:00:32.532315 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0330 16:00:32.532327 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0330 16:00:32.532341 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.632653 | |
| I0330 16:00:32.532354 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.40489 (* 0.3 = 0.721468 loss) | |
| I0330 16:00:32.532367 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.697895 (* 0.3 = 0.209368 loss) | |
| I0330 16:00:32.532380 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.612245 | |
| I0330 16:00:32.532392 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0330 16:00:32.532404 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.755102 | |
| I0330 16:00:32.532418 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.78039 (* 1 = 1.78039 loss) | |
| I0330 16:00:32.532433 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.511893 (* 1 = 0.511893 loss) | |
| I0330 16:00:32.532444 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0330 16:00:32.532456 13762 solver.cpp:245] Train net output #16: total_confidence = 0.147799 | |
| I0330 16:00:32.532469 13762 sgd_solver.cpp:106] Iteration 74500, lr = 0.01 | |
| I0330 16:02:46.909948 13762 solver.cpp:338] Iteration 75000, Testing net (#0) | |
| I0330 16:03:16.804090 13762 solver.cpp:393] Test loss: 2.21339 | |
| I0330 16:03:16.804136 13762 solver.cpp:406] Test net output #0: loss1/accuracy = 0.522433 | |
| I0330 16:03:16.804152 13762 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.877002 | |
| I0330 16:03:16.804168 13762 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.797144 | |
| I0330 16:03:16.804184 13762 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.67097 (* 0.3 = 0.50129 loss) | |
| I0330 16:03:16.804198 13762 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.437009 (* 0.3 = 0.131103 loss) | |
| I0330 16:03:16.804211 13762 solver.cpp:406] Test net output #5: loss2/accuracy = 0.69158 | |
| I0330 16:03:16.804224 13762 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.918819 | |
| I0330 16:03:16.804234 13762 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.882825 | |
| I0330 16:03:16.804249 13762 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.11888 (* 0.3 = 0.335664 loss) | |
| I0330 16:03:16.804262 13762 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.30017 (* 0.3 = 0.0900509 loss) | |
| I0330 16:03:16.804275 13762 solver.cpp:406] Test net output #10: loss3/accuracy = 0.779838 | |
| I0330 16:03:16.804286 13762 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.944363 | |
| I0330 16:03:16.804297 13762 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.894781 | |
| I0330 16:03:16.804311 13762 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.916041 (* 1 = 0.916041 loss) | |
| I0330 16:03:16.804325 13762 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.239239 (* 1 = 0.239239 loss) | |
| I0330 16:03:16.804337 13762 solver.cpp:406] Test net output #15: total_accuracy = 0.473 | |
| I0330 16:03:16.804348 13762 solver.cpp:406] Test net output #16: total_confidence = 0.447159 | |
| I0330 16:03:16.955721 13762 solver.cpp:229] Iteration 75000, loss = 2.55128 | |
| I0330 16:03:16.955824 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.536585 | |
| I0330 16:03:16.955842 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0330 16:03:16.955855 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.829268 | |
| I0330 16:03:16.955870 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.45588 (* 0.3 = 0.436763 loss) | |
| I0330 16:03:16.955885 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.47643 (* 0.3 = 0.142929 loss) | |
| I0330 16:03:16.955898 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.707317 | |
| I0330 16:03:16.955910 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 16:03:16.955922 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.829268 | |
| I0330 16:03:16.955935 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.974418 (* 0.3 = 0.292325 loss) | |
| I0330 16:03:16.955950 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.379256 (* 0.3 = 0.113777 loss) | |
| I0330 16:03:16.955962 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.902439 | |
| I0330 16:03:16.955974 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0330 16:03:16.955986 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97561 | |
| I0330 16:03:16.956001 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.338234 (* 1 = 0.338234 loss) | |
| I0330 16:03:16.956014 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.164282 (* 1 = 0.164282 loss) | |
| I0330 16:03:16.956027 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0330 16:03:16.956038 13762 solver.cpp:245] Train net output #16: total_confidence = 0.47087 | |
| I0330 16:03:16.956051 13762 sgd_solver.cpp:106] Iteration 75000, lr = 0.01 | |
| I0330 16:06:46.999320 13762 solver.cpp:229] Iteration 75500, loss = 2.51743 | |
| I0330 16:06:46.999480 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.444444 | |
| I0330 16:06:46.999500 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0330 16:06:46.999513 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.611111 | |
| I0330 16:06:46.999537 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06785 (* 0.3 = 0.620354 loss) | |
| I0330 16:06:46.999552 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.52326 (* 0.3 = 0.156978 loss) | |
| I0330 16:06:46.999565 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.638889 | |
| I0330 16:06:46.999578 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0330 16:06:46.999590 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.861111 | |
| I0330 16:06:46.999604 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.36019 (* 0.3 = 0.408057 loss) | |
| I0330 16:06:46.999619 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.418535 (* 0.3 = 0.125561 loss) | |
| I0330 16:06:46.999631 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0330 16:06:46.999644 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0330 16:06:46.999655 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0330 16:06:46.999670 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.788496 (* 1 = 0.788496 loss) | |
| I0330 16:06:46.999685 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.202735 (* 1 = 0.202735 loss) | |
| I0330 16:06:46.999696 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 16:06:46.999709 13762 solver.cpp:245] Train net output #16: total_confidence = 0.328739 | |
| I0330 16:06:46.999722 13762 sgd_solver.cpp:106] Iteration 75500, lr = 0.01 | |
| I0330 16:08:56.037783 13762 solver.cpp:229] Iteration 76000, loss = 2.48255 | |
| I0330 16:08:56.037910 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.384615 | |
| I0330 16:08:56.037930 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0330 16:08:56.037943 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.673077 | |
| I0330 16:08:56.037960 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.03941 (* 0.3 = 0.611823 loss) | |
| I0330 16:08:56.037974 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.627142 (* 0.3 = 0.188142 loss) | |
| I0330 16:08:56.037986 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0330 16:08:56.037999 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0330 16:08:56.038012 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.807692 | |
| I0330 16:08:56.038025 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.5092 (* 0.3 = 0.452759 loss) | |
| I0330 16:08:56.038040 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.474353 (* 0.3 = 0.142306 loss) | |
| I0330 16:08:56.038053 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0330 16:08:56.038064 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0330 16:08:56.038076 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.961538 | |
| I0330 16:08:56.038090 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.642566 (* 1 = 0.642566 loss) | |
| I0330 16:08:56.038105 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.19519 (* 1 = 0.19519 loss) | |
| I0330 16:08:56.038117 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0330 16:08:56.038130 13762 solver.cpp:245] Train net output #16: total_confidence = 0.273497 | |
| I0330 16:08:56.038141 13762 sgd_solver.cpp:106] Iteration 76000, lr = 0.01 | |
| I0330 16:11:05.244459 13762 solver.cpp:229] Iteration 76500, loss = 2.51794 | |
| I0330 16:11:05.244609 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.470588 | |
| I0330 16:11:05.244629 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0330 16:11:05.244642 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72549 | |
| I0330 16:11:05.244658 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73682 (* 0.3 = 0.521045 loss) | |
| I0330 16:11:05.244673 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.532344 (* 0.3 = 0.159703 loss) | |
| I0330 16:11:05.244686 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54902 | |
| I0330 16:11:05.244699 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 16:11:05.244711 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.823529 | |
| I0330 16:11:05.244729 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32103 (* 0.3 = 0.396309 loss) | |
| I0330 16:11:05.244743 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.416331 (* 0.3 = 0.124899 loss) | |
| I0330 16:11:05.244755 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.784314 | |
| I0330 16:11:05.244767 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0330 16:11:05.244779 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.941176 | |
| I0330 16:11:05.244793 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.716673 (* 1 = 0.716673 loss) | |
| I0330 16:11:05.244807 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.217739 (* 1 = 0.217739 loss) | |
| I0330 16:11:05.244820 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 16:11:05.244832 13762 solver.cpp:245] Train net output #16: total_confidence = 0.414619 | |
| I0330 16:11:05.244845 13762 sgd_solver.cpp:106] Iteration 76500, lr = 0.01 | |
| I0330 16:13:14.333773 13762 solver.cpp:229] Iteration 77000, loss = 2.4908 | |
| I0330 16:13:14.333884 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.565217 | |
| I0330 16:13:14.333904 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0330 16:13:14.333916 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.847826 | |
| I0330 16:13:14.333932 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.61292 (* 0.3 = 0.483877 loss) | |
| I0330 16:13:14.333947 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.480366 (* 0.3 = 0.14411 loss) | |
| I0330 16:13:14.333961 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.73913 | |
| I0330 16:13:14.333973 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.909091 | |
| I0330 16:13:14.333986 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.913043 | |
| I0330 16:13:14.334000 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.939467 (* 0.3 = 0.28184 loss) | |
| I0330 16:13:14.334014 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.296109 (* 0.3 = 0.0888326 loss) | |
| I0330 16:13:14.334028 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.956522 | |
| I0330 16:13:14.334039 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0330 16:13:14.334051 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0330 16:13:14.334076 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.162316 (* 1 = 0.162316 loss) | |
| I0330 16:13:14.334091 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0564361 (* 1 = 0.0564361 loss) | |
| I0330 16:13:14.334105 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0330 16:13:14.334115 13762 solver.cpp:245] Train net output #16: total_confidence = 0.483011 | |
| I0330 16:13:14.334128 13762 sgd_solver.cpp:106] Iteration 77000, lr = 0.01 | |
| I0330 16:15:23.626332 13762 solver.cpp:229] Iteration 77500, loss = 2.55586 | |
| I0330 16:15:23.626488 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.363636 | |
| I0330 16:15:23.626508 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0330 16:15:23.626523 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0330 16:15:23.626538 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.98249 (* 0.3 = 0.594746 loss) | |
| I0330 16:15:23.626554 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.585848 (* 0.3 = 0.175754 loss) | |
| I0330 16:15:23.626575 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0330 16:15:23.626588 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0330 16:15:23.626600 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0330 16:15:23.626616 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75072 (* 0.3 = 0.525217 loss) | |
| I0330 16:15:23.626629 13762 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.505183 (* 0.3 = 0.151555 loss) | |
| I0330 16:15:23.626642 13762 solver.cpp:245] Train net output #10: loss3/accuracy = 0.818182 | |
| I0330 16:15:23.626654 13762 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0330 16:15:23.626667 13762 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0330 16:15:23.626680 13762 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.96865 (* 1 = 0.96865 loss) | |
| I0330 16:15:23.626694 13762 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.280234 (* 1 = 0.280234 loss) | |
| I0330 16:15:23.626708 13762 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0330 16:15:23.626719 13762 solver.cpp:245] Train net output #16: total_confidence = 0.426039 | |
| I0330 16:15:23.626731 13762 sgd_solver.cpp:106] Iteration 77500, lr = 0.01 | |
| I0330 16:17:32.592768 13762 solver.cpp:229] Iteration 78000, loss = 2.54362 | |
| I0330 16:17:32.592943 13762 solver.cpp:245] Train net output #0: loss1/accuracy = 0.423077 | |
| I0330 16:17:32.592964 13762 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0330 16:17:32.592978 13762 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.634615 | |
| I0330 16:17:32.592995 13762 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.95772 (* 0.3 = 0.587317 loss) | |
| I0330 16:17:32.593011 13762 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.600092 (* 0.3 = 0.180028 loss) | |
| I0330 16:17:32.593024 13762 solver.cpp:245] Train net output #5: loss2/accuracy = 0.557692 | |
| I0330 16:17:32.593036 13762 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0330 16:17:32.593050 13762 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.807692 | |
| I0330 16:17:32.593063 13762 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.62843 (* 0.3 = 0.488529 loss) | |
| I0330 16:17:32.593078 13762 solver.cpp:245] Train net output #9: loss2/cross_entr |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment