Last active
November 12, 2016 21:12
-
-
Save Luonic/70c5f56b61dcf9cd75dd457a4fc0a2ed to your computer and use it in GitHub Desktop.
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
| I1112 07:17:41.890250 17193 solver.cpp:337] Iteration 0, Testing net (#0) | |
| I1112 07:17:41.896697 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 07:17:41.901717 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 07:17:49.161095 17193 solver.cpp:404] Test net output #0: accuracy = 0.02056 | |
| I1112 07:17:49.283001 17193 solver.cpp:228] Iteration 0, loss = 5.03651 | |
| I1112 07:17:49.283051 17193 solver.cpp:244] Train net output #0: loss = 5.03651 (* 1 = 5.03651 loss) | |
| I1112 07:17:49.283078 17193 sgd_solver.cpp:106] Iteration 0, lr = 0.01 | |
| I1112 07:19:59.706967 17193 solver.cpp:228] Iteration 1000, loss = 1.17764 | |
| I1112 07:19:59.707039 17193 solver.cpp:244] Train net output #0: loss = 1.17765 (* 1 = 1.17765 loss) | |
| I1112 07:19:59.707053 17193 sgd_solver.cpp:106] Iteration 1000, lr = 0.000187328 | |
| I1112 07:22:10.137840 17193 solver.cpp:228] Iteration 2000, loss = 0.696419 | |
| I1112 07:22:10.138025 17193 solver.cpp:244] Train net output #0: loss = 0.69642 (* 1 = 0.69642 loss) | |
| I1112 07:22:10.138043 17193 sgd_solver.cpp:106] Iteration 2000, lr = 0.000111594 | |
| I1112 07:24:20.554947 17193 solver.cpp:228] Iteration 3000, loss = 0.669033 | |
| I1112 07:24:20.555074 17193 solver.cpp:244] Train net output #0: loss = 0.669034 (* 1 = 0.669034 loss) | |
| I1112 07:24:20.555091 17193 sgd_solver.cpp:106] Iteration 3000, lr = 8.23842e-05 | |
| I1112 07:26:30.972342 17193 solver.cpp:228] Iteration 4000, loss = 0.416319 | |
| I1112 07:26:30.972483 17193 solver.cpp:244] Train net output #0: loss = 0.41632 (* 1 = 0.41632 loss) | |
| I1112 07:26:30.972494 17193 sgd_solver.cpp:106] Iteration 4000, lr = 6.64164e-05 | |
| I1112 07:28:41.274926 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_5000.caffemodel | |
| I1112 07:28:50.852340 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_5000.solverstate | |
| I1112 07:28:51.019891 17193 solver.cpp:337] Iteration 5000, Testing net (#0) | |
| I1112 07:28:51.019948 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 07:28:51.019956 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 07:28:58.222630 17193 solver.cpp:404] Test net output #0: accuracy = 0.93048 | |
| I1112 07:28:58.259209 17193 solver.cpp:228] Iteration 5000, loss = 0.322317 | |
| I1112 07:28:58.259263 17193 solver.cpp:244] Train net output #0: loss = 0.322318 (* 1 = 0.322318 loss) | |
| I1112 07:28:58.259276 17193 sgd_solver.cpp:106] Iteration 5000, lr = 5.6192e-05 | |
| I1112 07:31:08.661331 17193 solver.cpp:228] Iteration 6000, loss = 0.19025 | |
| I1112 07:31:08.661449 17193 solver.cpp:244] Train net output #0: loss = 0.190251 (* 1 = 0.190251 loss) | |
| I1112 07:31:08.661465 17193 sgd_solver.cpp:106] Iteration 6000, lr = 4.90165e-05 | |
| I1112 07:33:19.080974 17193 solver.cpp:228] Iteration 7000, loss = 0.164357 | |
| I1112 07:33:19.081104 17193 solver.cpp:244] Train net output #0: loss = 0.164358 (* 1 = 0.164358 loss) | |
| I1112 07:33:19.081120 17193 sgd_solver.cpp:106] Iteration 7000, lr = 4.36688e-05 | |
| I1112 07:35:29.481665 17193 solver.cpp:228] Iteration 8000, loss = 0.133433 | |
| I1112 07:35:29.481739 17193 solver.cpp:244] Train net output #0: loss = 0.133434 (* 1 = 0.133434 loss) | |
| I1112 07:35:29.481753 17193 sgd_solver.cpp:106] Iteration 8000, lr = 3.951e-05 | |
| I1112 07:37:39.918493 17193 solver.cpp:228] Iteration 9000, loss = 0.132508 | |
| I1112 07:37:39.918571 17193 solver.cpp:244] Train net output #0: loss = 0.132509 (* 1 = 0.132509 loss) | |
| I1112 07:37:39.918586 17193 sgd_solver.cpp:106] Iteration 9000, lr = 3.61714e-05 | |
| I1112 07:39:50.224310 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_10000.caffemodel | |
| I1112 07:40:05.595932 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_10000.solverstate | |
| I1112 07:40:05.745854 17193 solver.cpp:337] Iteration 10000, Testing net (#0) | |
| I1112 07:40:05.745901 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 07:40:05.745934 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 07:40:12.960891 17193 solver.cpp:404] Test net output #0: accuracy = 0.97712 | |
| I1112 07:40:12.995628 17193 solver.cpp:228] Iteration 10000, loss = 0.229462 | |
| I1112 07:40:12.995678 17193 solver.cpp:244] Train net output #0: loss = 0.229463 (* 1 = 0.229463 loss) | |
| I1112 07:40:12.995694 17193 sgd_solver.cpp:106] Iteration 10000, lr = 3.34245e-05 | |
| I1112 07:42:23.409602 17193 solver.cpp:228] Iteration 11000, loss = 0.206915 | |
| I1112 07:42:23.409744 17193 solver.cpp:244] Train net output #0: loss = 0.206916 (* 1 = 0.206916 loss) | |
| I1112 07:42:23.409757 17193 sgd_solver.cpp:106] Iteration 11000, lr = 3.11197e-05 | |
| I1112 07:44:33.804935 17193 solver.cpp:228] Iteration 12000, loss = 0.103708 | |
| I1112 07:44:33.805079 17193 solver.cpp:244] Train net output #0: loss = 0.10371 (* 1 = 0.10371 loss) | |
| I1112 07:44:33.805093 17193 sgd_solver.cpp:106] Iteration 12000, lr = 2.91545e-05 | |
| I1112 07:46:44.204833 17193 solver.cpp:228] Iteration 13000, loss = 0.0349696 | |
| I1112 07:46:44.204955 17193 solver.cpp:244] Train net output #0: loss = 0.0349711 (* 1 = 0.0349711 loss) | |
| I1112 07:46:44.204970 17193 sgd_solver.cpp:106] Iteration 13000, lr = 2.74565e-05 | |
| I1112 07:48:54.609351 17193 solver.cpp:228] Iteration 14000, loss = 0.0794354 | |
| I1112 07:48:54.609424 17193 solver.cpp:244] Train net output #0: loss = 0.079437 (* 1 = 0.079437 loss) | |
| I1112 07:48:54.609436 17193 sgd_solver.cpp:106] Iteration 14000, lr = 2.59726e-05 | |
| I1112 07:51:04.876116 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_15000.caffemodel | |
| I1112 07:51:05.437881 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_15000.solverstate | |
| I1112 07:51:06.640655 17193 solver.cpp:337] Iteration 15000, Testing net (#0) | |
| I1112 07:51:06.640708 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 07:51:06.640717 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 07:51:13.838026 17193 solver.cpp:404] Test net output #0: accuracy = 0.98464 | |
| I1112 07:51:13.874845 17193 solver.cpp:228] Iteration 15000, loss = 0.110722 | |
| I1112 07:51:13.874896 17193 solver.cpp:244] Train net output #0: loss = 0.110723 (* 1 = 0.110723 loss) | |
| I1112 07:51:13.874910 17193 sgd_solver.cpp:106] Iteration 15000, lr = 2.46633e-05 | |
| I1112 07:53:24.298288 17193 solver.cpp:228] Iteration 16000, loss = 0.126085 | |
| I1112 07:53:24.298444 17193 solver.cpp:244] Train net output #0: loss = 0.126087 (* 1 = 0.126087 loss) | |
| I1112 07:53:24.298460 17193 sgd_solver.cpp:106] Iteration 16000, lr = 2.34983e-05 | |
| I1112 07:55:34.711755 17193 solver.cpp:228] Iteration 17000, loss = 0.0510434 | |
| I1112 07:55:34.711902 17193 solver.cpp:244] Train net output #0: loss = 0.051045 (* 1 = 0.051045 loss) | |
| I1112 07:55:34.711917 17193 sgd_solver.cpp:106] Iteration 17000, lr = 2.24541e-05 | |
| I1112 07:57:45.143920 17193 solver.cpp:228] Iteration 18000, loss = 0.0825318 | |
| I1112 07:57:45.143995 17193 solver.cpp:244] Train net output #0: loss = 0.0825334 (* 1 = 0.0825334 loss) | |
| I1112 07:57:45.144006 17193 sgd_solver.cpp:106] Iteration 18000, lr = 2.15121e-05 | |
| I1112 07:59:55.565953 17193 solver.cpp:228] Iteration 19000, loss = 0.0832459 | |
| I1112 07:59:55.566093 17193 solver.cpp:244] Train net output #0: loss = 0.0832476 (* 1 = 0.0832476 loss) | |
| I1112 07:59:55.566107 17193 sgd_solver.cpp:106] Iteration 19000, lr = 2.06574e-05 | |
| I1112 08:02:05.872259 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_20000.caffemodel | |
| I1112 08:02:06.349077 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_20000.solverstate | |
| I1112 08:02:06.491251 17193 solver.cpp:337] Iteration 20000, Testing net (#0) | |
| I1112 08:02:06.491304 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:02:06.491312 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:02:13.691721 17193 solver.cpp:404] Test net output #0: accuracy = 0.98928 | |
| I1112 08:02:13.726101 17193 solver.cpp:228] Iteration 20000, loss = 0.0242369 | |
| I1112 08:02:13.726150 17193 solver.cpp:244] Train net output #0: loss = 0.0242386 (* 1 = 0.0242386 loss) | |
| I1112 08:02:13.726163 17193 sgd_solver.cpp:106] Iteration 20000, lr = 1.9878e-05 | |
| I1112 08:04:24.125469 17193 solver.cpp:228] Iteration 21000, loss = 0.0179401 | |
| I1112 08:04:24.125613 17193 solver.cpp:244] Train net output #0: loss = 0.0179418 (* 1 = 0.0179418 loss) | |
| I1112 08:04:24.125627 17193 sgd_solver.cpp:106] Iteration 21000, lr = 1.9164e-05 | |
| I1112 08:06:34.542148 17193 solver.cpp:228] Iteration 22000, loss = 0.0367586 | |
| I1112 08:06:34.542224 17193 solver.cpp:244] Train net output #0: loss = 0.0367604 (* 1 = 0.0367604 loss) | |
| I1112 08:06:34.542237 17193 sgd_solver.cpp:106] Iteration 22000, lr = 1.8507e-05 | |
| I1112 08:08:44.938272 17193 solver.cpp:228] Iteration 23000, loss = 0.0225611 | |
| I1112 08:08:44.938367 17193 solver.cpp:244] Train net output #0: loss = 0.0225629 (* 1 = 0.0225629 loss) | |
| I1112 08:08:44.938380 17193 sgd_solver.cpp:106] Iteration 23000, lr = 1.79003e-05 | |
| I1112 08:10:55.337188 17193 solver.cpp:228] Iteration 24000, loss = 0.0416818 | |
| I1112 08:10:55.337265 17193 solver.cpp:244] Train net output #0: loss = 0.0416836 (* 1 = 0.0416836 loss) | |
| I1112 08:10:55.337276 17193 sgd_solver.cpp:106] Iteration 24000, lr = 1.73381e-05 | |
| I1112 08:13:05.625404 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_25000.caffemodel | |
| I1112 08:13:06.125607 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_25000.solverstate | |
| I1112 08:13:06.645081 17193 solver.cpp:337] Iteration 25000, Testing net (#0) | |
| I1112 08:13:06.645134 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:13:06.645144 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:13:13.835492 17193 solver.cpp:404] Test net output #0: accuracy = 0.99328 | |
| I1112 08:13:13.869546 17193 solver.cpp:228] Iteration 25000, loss = 0.0595944 | |
| I1112 08:13:13.869596 17193 solver.cpp:244] Train net output #0: loss = 0.0595962 (* 1 = 0.0595962 loss) | |
| I1112 08:13:13.869609 17193 sgd_solver.cpp:106] Iteration 25000, lr = 1.68154e-05 | |
| I1112 08:15:24.286114 17193 solver.cpp:228] Iteration 26000, loss = 0.127027 | |
| I1112 08:15:24.286236 17193 solver.cpp:244] Train net output #0: loss = 0.127029 (* 1 = 0.127029 loss) | |
| I1112 08:15:24.286250 17193 sgd_solver.cpp:106] Iteration 26000, lr = 1.63281e-05 | |
| I1112 08:17:34.690212 17193 solver.cpp:228] Iteration 27000, loss = 0.00708207 | |
| I1112 08:17:34.690328 17193 solver.cpp:244] Train net output #0: loss = 0.00708402 (* 1 = 0.00708402 loss) | |
| I1112 08:17:34.690341 17193 sgd_solver.cpp:106] Iteration 27000, lr = 1.58725e-05 | |
| I1112 08:19:45.100325 17193 solver.cpp:228] Iteration 28000, loss = 0.025895 | |
| I1112 08:19:45.100450 17193 solver.cpp:244] Train net output #0: loss = 0.0258969 (* 1 = 0.0258969 loss) | |
| I1112 08:19:45.100466 17193 sgd_solver.cpp:106] Iteration 28000, lr = 1.54455e-05 | |
| I1112 08:21:55.505355 17193 solver.cpp:228] Iteration 29000, loss = 0.0164592 | |
| I1112 08:21:55.505441 17193 solver.cpp:244] Train net output #0: loss = 0.0164611 (* 1 = 0.0164611 loss) | |
| I1112 08:21:55.505453 17193 sgd_solver.cpp:106] Iteration 29000, lr = 1.50443e-05 | |
| I1112 08:24:05.777113 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_30000.caffemodel | |
| I1112 08:24:40.955708 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_30000.solverstate | |
| I1112 08:24:41.119065 17193 solver.cpp:337] Iteration 30000, Testing net (#0) | |
| I1112 08:24:41.119114 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:24:41.119134 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:24:48.340046 17193 solver.cpp:404] Test net output #0: accuracy = 0.9932 | |
| I1112 08:24:48.374820 17193 solver.cpp:228] Iteration 30000, loss = 0.0249928 | |
| I1112 08:24:48.374873 17193 solver.cpp:244] Train net output #0: loss = 0.0249948 (* 1 = 0.0249948 loss) | |
| I1112 08:24:48.374888 17193 sgd_solver.cpp:106] Iteration 30000, lr = 1.46667e-05 | |
| I1112 08:26:58.790920 17193 solver.cpp:228] Iteration 31000, loss = 0.160548 | |
| I1112 08:26:58.790997 17193 solver.cpp:244] Train net output #0: loss = 0.16055 (* 1 = 0.16055 loss) | |
| I1112 08:26:58.791012 17193 sgd_solver.cpp:106] Iteration 31000, lr = 1.43105e-05 | |
| I1112 08:29:09.200445 17193 solver.cpp:228] Iteration 32000, loss = 0.0215395 | |
| I1112 08:29:09.200536 17193 solver.cpp:244] Train net output #0: loss = 0.0215413 (* 1 = 0.0215413 loss) | |
| I1112 08:29:09.200549 17193 sgd_solver.cpp:106] Iteration 32000, lr = 1.39738e-05 | |
| I1112 08:31:19.602881 17193 solver.cpp:228] Iteration 33000, loss = 0.0146531 | |
| I1112 08:31:19.603001 17193 solver.cpp:244] Train net output #0: loss = 0.014655 (* 1 = 0.014655 loss) | |
| I1112 08:31:19.603016 17193 sgd_solver.cpp:106] Iteration 33000, lr = 1.3655e-05 | |
| I1112 08:33:29.995241 17193 solver.cpp:228] Iteration 34000, loss = 0.0363096 | |
| I1112 08:33:29.995329 17193 solver.cpp:244] Train net output #0: loss = 0.0363116 (* 1 = 0.0363116 loss) | |
| I1112 08:33:29.995344 17193 sgd_solver.cpp:106] Iteration 34000, lr = 1.33527e-05 | |
| I1112 08:35:40.281373 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_35000.caffemodel | |
| I1112 08:35:41.298012 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_35000.solverstate | |
| I1112 08:35:41.493650 17193 solver.cpp:337] Iteration 35000, Testing net (#0) | |
| I1112 08:35:41.493703 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:35:41.493712 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:35:48.688002 17193 solver.cpp:404] Test net output #0: accuracy = 0.99472 | |
| I1112 08:35:48.722436 17193 solver.cpp:228] Iteration 35000, loss = 0.0176381 | |
| I1112 08:35:48.722486 17193 solver.cpp:244] Train net output #0: loss = 0.0176401 (* 1 = 0.0176401 loss) | |
| I1112 08:35:48.722501 17193 sgd_solver.cpp:106] Iteration 35000, lr = 1.30656e-05 | |
| I1112 08:37:59.133245 17193 solver.cpp:228] Iteration 36000, loss = 0.0154841 | |
| I1112 08:37:59.133317 17193 solver.cpp:244] Train net output #0: loss = 0.015486 (* 1 = 0.015486 loss) | |
| I1112 08:37:59.133333 17193 sgd_solver.cpp:106] Iteration 36000, lr = 1.27925e-05 | |
| I1112 08:40:09.553462 17193 solver.cpp:228] Iteration 37000, loss = 0.0470804 | |
| I1112 08:40:09.553549 17193 solver.cpp:244] Train net output #0: loss = 0.0470821 (* 1 = 0.0470821 loss) | |
| I1112 08:40:09.553565 17193 sgd_solver.cpp:106] Iteration 37000, lr = 1.25323e-05 | |
| I1112 08:42:19.991109 17193 solver.cpp:228] Iteration 38000, loss = 0.0206935 | |
| I1112 08:42:19.991261 17193 solver.cpp:244] Train net output #0: loss = 0.0206952 (* 1 = 0.0206952 loss) | |
| I1112 08:42:19.991274 17193 sgd_solver.cpp:106] Iteration 38000, lr = 1.22842e-05 | |
| I1112 08:44:30.420471 17193 solver.cpp:228] Iteration 39000, loss = 0.0118376 | |
| I1112 08:44:30.420563 17193 solver.cpp:244] Train net output #0: loss = 0.0118393 (* 1 = 0.0118393 loss) | |
| I1112 08:44:30.420578 17193 sgd_solver.cpp:106] Iteration 39000, lr = 1.20472e-05 | |
| I1112 08:46:40.680351 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_40000.caffemodel | |
| I1112 08:47:05.264895 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_40000.solverstate | |
| I1112 08:47:05.427218 17193 solver.cpp:337] Iteration 40000, Testing net (#0) | |
| I1112 08:47:05.427274 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:47:05.427284 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:47:12.628058 17193 solver.cpp:404] Test net output #0: accuracy = 0.99584 | |
| I1112 08:47:12.665499 17193 solver.cpp:228] Iteration 40000, loss = 0.0695327 | |
| I1112 08:47:12.665550 17193 solver.cpp:244] Train net output #0: loss = 0.0695344 (* 1 = 0.0695344 loss) | |
| I1112 08:47:12.665565 17193 sgd_solver.cpp:106] Iteration 40000, lr = 1.18207e-05 | |
| I1112 08:49:23.098924 17193 solver.cpp:228] Iteration 41000, loss = 0.0326135 | |
| I1112 08:49:23.099020 17193 solver.cpp:244] Train net output #0: loss = 0.0326152 (* 1 = 0.0326152 loss) | |
| I1112 08:49:23.099032 17193 sgd_solver.cpp:106] Iteration 41000, lr = 1.16038e-05 | |
| I1112 08:51:33.504154 17193 solver.cpp:228] Iteration 42000, loss = 0.0406619 | |
| I1112 08:51:33.504254 17193 solver.cpp:244] Train net output #0: loss = 0.0406636 (* 1 = 0.0406636 loss) | |
| I1112 08:51:33.504267 17193 sgd_solver.cpp:106] Iteration 42000, lr = 1.1396e-05 | |
| I1112 08:53:43.923717 17193 solver.cpp:228] Iteration 43000, loss = 0.0245159 | |
| I1112 08:53:43.923861 17193 solver.cpp:244] Train net output #0: loss = 0.0245178 (* 1 = 0.0245178 loss) | |
| I1112 08:53:43.923876 17193 sgd_solver.cpp:106] Iteration 43000, lr = 1.11967e-05 | |
| I1112 08:55:54.325821 17193 solver.cpp:228] Iteration 44000, loss = 0.0107809 | |
| I1112 08:55:54.325909 17193 solver.cpp:244] Train net output #0: loss = 0.0107828 (* 1 = 0.0107828 loss) | |
| I1112 08:55:54.325924 17193 sgd_solver.cpp:106] Iteration 44000, lr = 1.10053e-05 | |
| I1112 08:58:04.604471 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_45000.caffemodel | |
| I1112 08:58:35.298804 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_45000.solverstate | |
| I1112 08:58:35.444377 17193 solver.cpp:337] Iteration 45000, Testing net (#0) | |
| I1112 08:58:35.444430 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 08:58:35.444442 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 08:58:42.636595 17193 solver.cpp:404] Test net output #0: accuracy = 0.99536 | |
| I1112 08:58:42.670956 17193 solver.cpp:228] Iteration 45000, loss = 0.0319046 | |
| I1112 08:58:42.671005 17193 solver.cpp:244] Train net output #0: loss = 0.0319065 (* 1 = 0.0319065 loss) | |
| I1112 08:58:42.671020 17193 sgd_solver.cpp:106] Iteration 45000, lr = 1.08214e-05 | |
| I1112 09:00:53.093411 17193 solver.cpp:228] Iteration 46000, loss = 0.00533195 | |
| I1112 09:00:53.093483 17193 solver.cpp:244] Train net output #0: loss = 0.0053338 (* 1 = 0.0053338 loss) | |
| I1112 09:00:53.093494 17193 sgd_solver.cpp:106] Iteration 46000, lr = 1.06445e-05 | |
| I1112 09:03:03.509090 17193 solver.cpp:228] Iteration 47000, loss = 0.0167692 | |
| I1112 09:03:03.509167 17193 solver.cpp:244] Train net output #0: loss = 0.016771 (* 1 = 0.016771 loss) | |
| I1112 09:03:03.509181 17193 sgd_solver.cpp:106] Iteration 47000, lr = 1.04742e-05 | |
| I1112 09:05:13.914008 17193 solver.cpp:228] Iteration 48000, loss = 0.0141988 | |
| I1112 09:05:13.914146 17193 solver.cpp:244] Train net output #0: loss = 0.0142006 (* 1 = 0.0142006 loss) | |
| I1112 09:05:13.914158 17193 sgd_solver.cpp:106] Iteration 48000, lr = 1.03101e-05 | |
| I1112 09:07:24.325129 17193 solver.cpp:228] Iteration 49000, loss = 0.0368493 | |
| I1112 09:07:24.325279 17193 solver.cpp:244] Train net output #0: loss = 0.0368512 (* 1 = 0.0368512 loss) | |
| I1112 09:07:24.325294 17193 sgd_solver.cpp:106] Iteration 49000, lr = 1.01519e-05 | |
| I1112 09:09:34.616418 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_50000.caffemodel | |
| I1112 09:10:08.319231 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_50000.solverstate | |
| I1112 09:10:08.465332 17193 solver.cpp:337] Iteration 50000, Testing net (#0) | |
| I1112 09:10:08.465378 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 09:10:08.465399 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 09:10:15.664849 17193 solver.cpp:404] Test net output #0: accuracy = 0.9956 | |
| I1112 09:10:15.699282 17193 solver.cpp:228] Iteration 50000, loss = 0.0129046 | |
| I1112 09:10:15.699332 17193 solver.cpp:244] Train net output #0: loss = 0.0129065 (* 1 = 0.0129065 loss) | |
| I1112 09:10:15.699348 17193 sgd_solver.cpp:106] Iteration 50000, lr = 9.99925e-06 | |
| I1112 09:12:26.137650 17193 solver.cpp:228] Iteration 51000, loss = 0.0151948 | |
| I1112 09:12:26.137753 17193 solver.cpp:244] Train net output #0: loss = 0.0151966 (* 1 = 0.0151966 loss) | |
| I1112 09:12:26.137764 17193 sgd_solver.cpp:106] Iteration 51000, lr = 9.85185e-06 | |
| I1112 09:14:36.551502 17193 solver.cpp:228] Iteration 52000, loss = 0.00694739 | |
| I1112 09:14:36.551630 17193 solver.cpp:244] Train net output #0: loss = 0.00694925 (* 1 = 0.00694925 loss) | |
| I1112 09:14:36.551643 17193 sgd_solver.cpp:106] Iteration 52000, lr = 9.70943e-06 | |
| I1112 09:16:46.962292 17193 solver.cpp:228] Iteration 53000, loss = 0.0581361 | |
| I1112 09:16:46.962364 17193 solver.cpp:244] Train net output #0: loss = 0.058138 (* 1 = 0.058138 loss) | |
| I1112 09:16:46.962378 17193 sgd_solver.cpp:106] Iteration 53000, lr = 9.57172e-06 | |
| I1112 09:18:57.390095 17193 solver.cpp:228] Iteration 54000, loss = 0.00509106 | |
| I1112 09:18:57.390169 17193 solver.cpp:244] Train net output #0: loss = 0.00509299 (* 1 = 0.00509299 loss) | |
| I1112 09:18:57.390182 17193 sgd_solver.cpp:106] Iteration 54000, lr = 9.43848e-06 | |
| I1112 09:21:07.676455 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_55000.caffemodel | |
| I1112 09:21:41.568927 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_55000.solverstate | |
| I1112 09:21:41.707314 17193 solver.cpp:337] Iteration 55000, Testing net (#0) | |
| I1112 09:21:41.707366 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 09:21:41.707376 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 09:21:48.909824 17193 solver.cpp:404] Test net output #0: accuracy = 0.9968 | |
| I1112 09:21:48.944203 17193 solver.cpp:228] Iteration 55000, loss = 0.0143865 | |
| I1112 09:21:48.944253 17193 solver.cpp:244] Train net output #0: loss = 0.0143885 (* 1 = 0.0143885 loss) | |
| I1112 09:21:48.944269 17193 sgd_solver.cpp:106] Iteration 55000, lr = 9.30949e-06 | |
| I1112 09:23:59.372443 17193 solver.cpp:228] Iteration 56000, loss = 0.0490061 | |
| I1112 09:23:59.372586 17193 solver.cpp:244] Train net output #0: loss = 0.049008 (* 1 = 0.049008 loss) | |
| I1112 09:23:59.372598 17193 sgd_solver.cpp:106] Iteration 56000, lr = 9.18454e-06 | |
| I1112 09:26:09.836683 17193 solver.cpp:228] Iteration 57000, loss = 0.00615001 | |
| I1112 09:26:09.836791 17193 solver.cpp:244] Train net output #0: loss = 0.00615201 (* 1 = 0.00615201 loss) | |
| I1112 09:26:09.836814 17193 sgd_solver.cpp:106] Iteration 57000, lr = 9.06343e-06 | |
| I1112 09:28:20.250309 17193 solver.cpp:228] Iteration 58000, loss = 0.0483592 | |
| I1112 09:28:20.250403 17193 solver.cpp:244] Train net output #0: loss = 0.0483612 (* 1 = 0.0483612 loss) | |
| I1112 09:28:20.250414 17193 sgd_solver.cpp:106] Iteration 58000, lr = 8.94599e-06 | |
| I1112 09:30:30.651531 17193 solver.cpp:228] Iteration 59000, loss = 0.00149275 | |
| I1112 09:30:30.651623 17193 solver.cpp:244] Train net output #0: loss = 0.00149466 (* 1 = 0.00149466 loss) | |
| I1112 09:30:30.651633 17193 sgd_solver.cpp:106] Iteration 59000, lr = 8.83204e-06 | |
| I1112 09:32:41.297025 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_60000.caffemodel | |
| I1112 09:32:57.334919 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_60000.solverstate | |
| I1112 09:32:57.475343 17193 solver.cpp:337] Iteration 60000, Testing net (#0) | |
| I1112 09:32:57.475392 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 09:32:57.475399 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 09:33:04.681656 17193 solver.cpp:404] Test net output #0: accuracy = 0.99696 | |
| I1112 09:33:04.717983 17193 solver.cpp:228] Iteration 60000, loss = 0.0148324 | |
| I1112 09:33:04.718032 17193 solver.cpp:244] Train net output #0: loss = 0.0148343 (* 1 = 0.0148343 loss) | |
| I1112 09:33:04.718049 17193 sgd_solver.cpp:106] Iteration 60000, lr = 8.72141e-06 | |
| I1112 09:35:15.128329 17193 solver.cpp:228] Iteration 61000, loss = 0.0273681 | |
| I1112 09:35:15.128435 17193 solver.cpp:244] Train net output #0: loss = 0.0273699 (* 1 = 0.0273699 loss) | |
| I1112 09:35:15.128448 17193 sgd_solver.cpp:106] Iteration 61000, lr = 8.61397e-06 | |
| I1112 09:37:25.563390 17193 solver.cpp:228] Iteration 62000, loss = 0.01563 | |
| I1112 09:37:25.563499 17193 solver.cpp:244] Train net output #0: loss = 0.0156319 (* 1 = 0.0156319 loss) | |
| I1112 09:37:25.563514 17193 sgd_solver.cpp:106] Iteration 62000, lr = 8.50957e-06 | |
| I1112 09:39:36.000689 17193 solver.cpp:228] Iteration 63000, loss = 0.00251313 | |
| I1112 09:39:36.000777 17193 solver.cpp:244] Train net output #0: loss = 0.002515 (* 1 = 0.002515 loss) | |
| I1112 09:39:36.000792 17193 sgd_solver.cpp:106] Iteration 63000, lr = 8.40807e-06 | |
| I1112 09:41:46.383656 17193 solver.cpp:228] Iteration 64000, loss = 0.019572 | |
| I1112 09:41:46.383744 17193 solver.cpp:244] Train net output #0: loss = 0.019574 (* 1 = 0.019574 loss) | |
| I1112 09:41:46.383756 17193 sgd_solver.cpp:106] Iteration 64000, lr = 8.30935e-06 | |
| I1112 09:43:56.653594 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_65000.caffemodel | |
| I1112 09:44:10.757145 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_65000.solverstate | |
| I1112 09:44:10.898671 17193 solver.cpp:337] Iteration 65000, Testing net (#0) | |
| I1112 09:44:10.898712 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 09:44:10.898733 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 09:44:18.110404 17193 solver.cpp:404] Test net output #0: accuracy = 0.99688 | |
| I1112 09:44:18.146694 17193 solver.cpp:228] Iteration 65000, loss = 0.0310616 | |
| I1112 09:44:18.146747 17193 solver.cpp:244] Train net output #0: loss = 0.0310635 (* 1 = 0.0310635 loss) | |
| I1112 09:44:18.146764 17193 sgd_solver.cpp:106] Iteration 65000, lr = 8.21329e-06 | |
| I1112 09:46:28.552021 17193 solver.cpp:228] Iteration 66000, loss = 0.0165207 | |
| I1112 09:46:28.552111 17193 solver.cpp:244] Train net output #0: loss = 0.0165226 (* 1 = 0.0165226 loss) | |
| I1112 09:46:28.552124 17193 sgd_solver.cpp:106] Iteration 66000, lr = 8.11979e-06 | |
| I1112 09:48:38.951653 17193 solver.cpp:228] Iteration 67000, loss = 0.0124637 | |
| I1112 09:48:38.951774 17193 solver.cpp:244] Train net output #0: loss = 0.0124655 (* 1 = 0.0124655 loss) | |
| I1112 09:48:38.951791 17193 sgd_solver.cpp:106] Iteration 67000, lr = 8.02873e-06 | |
| I1112 09:50:49.349213 17193 solver.cpp:228] Iteration 68000, loss = 0.00470868 | |
| I1112 09:50:49.349287 17193 solver.cpp:244] Train net output #0: loss = 0.00471048 (* 1 = 0.00471048 loss) | |
| I1112 09:50:49.349298 17193 sgd_solver.cpp:106] Iteration 68000, lr = 7.94003e-06 | |
| I1112 09:52:59.769484 17193 solver.cpp:228] Iteration 69000, loss = 0.0139715 | |
| I1112 09:52:59.769562 17193 solver.cpp:244] Train net output #0: loss = 0.0139733 (* 1 = 0.0139733 loss) | |
| I1112 09:52:59.769575 17193 sgd_solver.cpp:106] Iteration 69000, lr = 7.85357e-06 | |
| I1112 09:55:10.071915 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_70000.caffemodel | |
| I1112 09:55:22.692672 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_70000.solverstate | |
| I1112 09:55:22.836462 17193 solver.cpp:337] Iteration 70000, Testing net (#0) | |
| I1112 09:55:22.836503 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 09:55:22.836524 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 09:55:30.034639 17193 solver.cpp:404] Test net output #0: accuracy = 0.99728 | |
| I1112 09:55:30.073004 17193 solver.cpp:228] Iteration 70000, loss = 0.108386 | |
| I1112 09:55:30.073046 17193 solver.cpp:244] Train net output #0: loss = 0.108388 (* 1 = 0.108388 loss) | |
| I1112 09:55:30.073072 17193 sgd_solver.cpp:106] Iteration 70000, lr = 7.76928e-06 | |
| I1112 09:57:40.505363 17193 solver.cpp:228] Iteration 71000, loss = 0.0419956 | |
| I1112 09:57:40.505523 17193 solver.cpp:244] Train net output #0: loss = 0.0419974 (* 1 = 0.0419974 loss) | |
| I1112 09:57:40.505538 17193 sgd_solver.cpp:106] Iteration 71000, lr = 7.68707e-06 | |
| I1112 09:59:50.922361 17193 solver.cpp:228] Iteration 72000, loss = 0.00742958 | |
| I1112 09:59:50.922487 17193 solver.cpp:244] Train net output #0: loss = 0.00743146 (* 1 = 0.00743146 loss) | |
| I1112 09:59:50.922502 17193 sgd_solver.cpp:106] Iteration 72000, lr = 7.60686e-06 | |
| I1112 10:02:01.345787 17193 solver.cpp:228] Iteration 73000, loss = 0.0119934 | |
| I1112 10:02:01.345859 17193 solver.cpp:244] Train net output #0: loss = 0.0119953 (* 1 = 0.0119953 loss) | |
| I1112 10:02:01.345870 17193 sgd_solver.cpp:106] Iteration 73000, lr = 7.52858e-06 | |
| I1112 10:04:11.761543 17193 solver.cpp:228] Iteration 74000, loss = 0.00303226 | |
| I1112 10:04:11.761659 17193 solver.cpp:244] Train net output #0: loss = 0.00303402 (* 1 = 0.00303402 loss) | |
| I1112 10:04:11.761672 17193 sgd_solver.cpp:106] Iteration 74000, lr = 7.45215e-06 | |
| I1112 10:06:22.030303 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_75000.caffemodel | |
| I1112 10:06:40.030457 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_75000.solverstate | |
| I1112 10:06:40.170608 17193 solver.cpp:337] Iteration 75000, Testing net (#0) | |
| I1112 10:06:40.170651 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 10:06:40.170672 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 10:06:47.364522 17193 solver.cpp:404] Test net output #0: accuracy = 0.99696 | |
| I1112 10:06:47.400594 17193 solver.cpp:228] Iteration 75000, loss = 0.00312457 | |
| I1112 10:06:47.400636 17193 solver.cpp:244] Train net output #0: loss = 0.00312638 (* 1 = 0.00312638 loss) | |
| I1112 10:06:47.400663 17193 sgd_solver.cpp:106] Iteration 75000, lr = 7.37751e-06 | |
| I1112 10:08:57.829016 17193 solver.cpp:228] Iteration 76000, loss = 0.0172165 | |
| I1112 10:08:57.829109 17193 solver.cpp:244] Train net output #0: loss = 0.0172183 (* 1 = 0.0172183 loss) | |
| I1112 10:08:57.829123 17193 sgd_solver.cpp:106] Iteration 76000, lr = 7.30459e-06 | |
| I1112 10:11:08.286813 17193 solver.cpp:228] Iteration 77000, loss = 0.0753643 | |
| I1112 10:11:08.286932 17193 solver.cpp:244] Train net output #0: loss = 0.075366 (* 1 = 0.075366 loss) | |
| I1112 10:11:08.286945 17193 sgd_solver.cpp:106] Iteration 77000, lr = 7.23333e-06 | |
| I1112 10:13:18.697801 17193 solver.cpp:228] Iteration 78000, loss = 0.0263328 | |
| I1112 10:13:18.697918 17193 solver.cpp:244] Train net output #0: loss = 0.0263345 (* 1 = 0.0263345 loss) | |
| I1112 10:13:18.697932 17193 sgd_solver.cpp:106] Iteration 78000, lr = 7.16367e-06 | |
| I1112 10:15:29.110260 17193 solver.cpp:228] Iteration 79000, loss = 0.0326599 | |
| I1112 10:15:29.110332 17193 solver.cpp:244] Train net output #0: loss = 0.0326616 (* 1 = 0.0326616 loss) | |
| I1112 10:15:29.110342 17193 sgd_solver.cpp:106] Iteration 79000, lr = 7.09556e-06 | |
| I1112 10:17:39.377058 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_80000.caffemodel | |
| I1112 10:18:05.402990 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_80000.solverstate | |
| I1112 10:18:05.542412 17193 solver.cpp:337] Iteration 80000, Testing net (#0) | |
| I1112 10:18:05.542459 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 10:18:05.542469 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 10:18:12.776607 17193 solver.cpp:404] Test net output #0: accuracy = 0.99768 | |
| I1112 10:18:12.812860 17193 solver.cpp:228] Iteration 80000, loss = 0.00236453 | |
| I1112 10:18:12.812911 17193 solver.cpp:244] Train net output #0: loss = 0.00236628 (* 1 = 0.00236628 loss) | |
| I1112 10:18:12.812924 17193 sgd_solver.cpp:106] Iteration 80000, lr = 7.02894e-06 | |
| I1112 10:20:23.220958 17193 solver.cpp:228] Iteration 81000, loss = 0.00765635 | |
| I1112 10:20:23.221093 17193 solver.cpp:244] Train net output #0: loss = 0.00765813 (* 1 = 0.00765813 loss) | |
| I1112 10:20:23.221107 17193 sgd_solver.cpp:106] Iteration 81000, lr = 6.96376e-06 | |
| I1112 10:22:33.641763 17193 solver.cpp:228] Iteration 82000, loss = 0.0218123 | |
| I1112 10:22:33.641829 17193 solver.cpp:244] Train net output #0: loss = 0.0218141 (* 1 = 0.0218141 loss) | |
| I1112 10:22:33.641840 17193 sgd_solver.cpp:106] Iteration 82000, lr = 6.89997e-06 | |
| I1112 10:24:44.056349 17193 solver.cpp:228] Iteration 83000, loss = 0.00593235 | |
| I1112 10:24:44.056419 17193 solver.cpp:244] Train net output #0: loss = 0.00593417 (* 1 = 0.00593417 loss) | |
| I1112 10:24:44.056432 17193 sgd_solver.cpp:106] Iteration 83000, lr = 6.83753e-06 | |
| I1112 10:26:54.478200 17193 solver.cpp:228] Iteration 84000, loss = 0.0277179 | |
| I1112 10:26:54.478307 17193 solver.cpp:244] Train net output #0: loss = 0.0277197 (* 1 = 0.0277197 loss) | |
| I1112 10:26:54.478332 17193 sgd_solver.cpp:106] Iteration 84000, lr = 6.77639e-06 | |
| I1112 10:29:04.758725 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_85000.caffemodel | |
| I1112 10:29:34.652869 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_85000.solverstate | |
| I1112 10:29:34.792742 17193 solver.cpp:337] Iteration 85000, Testing net (#0) | |
| I1112 10:29:34.792881 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 10:29:34.792891 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 10:29:41.997611 17193 solver.cpp:404] Test net output #0: accuracy = 0.99736 | |
| I1112 10:29:42.033536 17193 solver.cpp:228] Iteration 85000, loss = 0.078367 | |
| I1112 10:29:42.033588 17193 solver.cpp:244] Train net output #0: loss = 0.0783688 (* 1 = 0.0783688 loss) | |
| I1112 10:29:42.033602 17193 sgd_solver.cpp:106] Iteration 85000, lr = 6.71652e-06 | |
| I1112 10:31:52.444777 17193 solver.cpp:228] Iteration 86000, loss = 0.00343338 | |
| I1112 10:31:52.444869 17193 solver.cpp:244] Train net output #0: loss = 0.00343519 (* 1 = 0.00343519 loss) | |
| I1112 10:31:52.444882 17193 sgd_solver.cpp:106] Iteration 86000, lr = 6.65786e-06 | |
| I1112 10:34:02.856685 17193 solver.cpp:228] Iteration 87000, loss = 0.00937355 | |
| I1112 10:34:02.856756 17193 solver.cpp:244] Train net output #0: loss = 0.00937536 (* 1 = 0.00937536 loss) | |
| I1112 10:34:02.856767 17193 sgd_solver.cpp:106] Iteration 87000, lr = 6.60039e-06 | |
| I1112 10:36:13.279165 17193 solver.cpp:228] Iteration 88000, loss = 0.00986386 | |
| I1112 10:36:13.279233 17193 solver.cpp:244] Train net output #0: loss = 0.00986569 (* 1 = 0.00986569 loss) | |
| I1112 10:36:13.279245 17193 sgd_solver.cpp:106] Iteration 88000, lr = 6.54406e-06 | |
| I1112 10:38:23.689203 17193 solver.cpp:228] Iteration 89000, loss = 0.0362355 | |
| I1112 10:38:23.689318 17193 solver.cpp:244] Train net output #0: loss = 0.0362374 (* 1 = 0.0362374 loss) | |
| I1112 10:38:23.689332 17193 sgd_solver.cpp:106] Iteration 89000, lr = 6.48883e-06 | |
| I1112 10:40:33.965747 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_90000.caffemodel | |
| I1112 10:40:53.562062 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_90000.solverstate | |
| I1112 10:40:53.702118 17193 solver.cpp:337] Iteration 90000, Testing net (#0) | |
| I1112 10:40:53.702167 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 10:40:53.702174 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 10:41:00.910392 17193 solver.cpp:404] Test net output #0: accuracy = 0.99584 | |
| I1112 10:41:00.946714 17193 solver.cpp:228] Iteration 90000, loss = 0.042136 | |
| I1112 10:41:00.946763 17193 solver.cpp:244] Train net output #0: loss = 0.0421379 (* 1 = 0.0421379 loss) | |
| I1112 10:41:00.946775 17193 sgd_solver.cpp:106] Iteration 90000, lr = 6.43469e-06 | |
| I1112 10:43:11.362624 17193 solver.cpp:228] Iteration 91000, loss = 0.00140887 | |
| I1112 10:43:11.362699 17193 solver.cpp:244] Train net output #0: loss = 0.00141071 (* 1 = 0.00141071 loss) | |
| I1112 10:43:11.362710 17193 sgd_solver.cpp:106] Iteration 91000, lr = 6.38159e-06 | |
| I1112 10:45:21.768270 17193 solver.cpp:228] Iteration 92000, loss = 0.00385397 | |
| I1112 10:45:21.768484 17193 solver.cpp:244] Train net output #0: loss = 0.0038558 (* 1 = 0.0038558 loss) | |
| I1112 10:45:21.768501 17193 sgd_solver.cpp:106] Iteration 92000, lr = 6.32949e-06 | |
| I1112 10:47:32.178804 17193 solver.cpp:228] Iteration 93000, loss = 0.00479621 | |
| I1112 10:47:32.178879 17193 solver.cpp:244] Train net output #0: loss = 0.00479804 (* 1 = 0.00479804 loss) | |
| I1112 10:47:32.178892 17193 sgd_solver.cpp:106] Iteration 93000, lr = 6.27838e-06 | |
| I1112 10:49:42.582052 17193 solver.cpp:228] Iteration 94000, loss = 0.00249722 | |
| I1112 10:49:42.582124 17193 solver.cpp:244] Train net output #0: loss = 0.00249908 (* 1 = 0.00249908 loss) | |
| I1112 10:49:42.582136 17193 sgd_solver.cpp:106] Iteration 94000, lr = 6.22822e-06 | |
| I1112 10:51:52.869441 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_95000.caffemodel | |
| I1112 10:52:22.938619 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_95000.solverstate | |
| I1112 10:52:23.077471 17193 solver.cpp:337] Iteration 95000, Testing net (#0) | |
| I1112 10:52:23.077520 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 10:52:23.077528 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 10:52:30.274521 17193 solver.cpp:404] Test net output #0: accuracy = 0.99776 | |
| I1112 10:52:30.310628 17193 solver.cpp:228] Iteration 95000, loss = 0.00168234 | |
| I1112 10:52:30.310679 17193 solver.cpp:244] Train net output #0: loss = 0.00168415 (* 1 = 0.00168415 loss) | |
| I1112 10:52:30.310694 17193 sgd_solver.cpp:106] Iteration 95000, lr = 6.17899e-06 | |
| I1112 10:54:40.724043 17193 solver.cpp:228] Iteration 96000, loss = 0.0257581 | |
| I1112 10:54:40.724165 17193 solver.cpp:244] Train net output #0: loss = 0.0257599 (* 1 = 0.0257599 loss) | |
| I1112 10:54:40.724179 17193 sgd_solver.cpp:106] Iteration 96000, lr = 6.13066e-06 | |
| I1112 10:56:51.138087 17193 solver.cpp:228] Iteration 97000, loss = 0.0371108 | |
| I1112 10:56:51.138206 17193 solver.cpp:244] Train net output #0: loss = 0.0371125 (* 1 = 0.0371125 loss) | |
| I1112 10:56:51.138219 17193 sgd_solver.cpp:106] Iteration 97000, lr = 6.0832e-06 | |
| I1112 10:59:01.554452 17193 solver.cpp:228] Iteration 98000, loss = 0.0573133 | |
| I1112 10:59:01.554525 17193 solver.cpp:244] Train net output #0: loss = 0.0573151 (* 1 = 0.0573151 loss) | |
| I1112 10:59:01.554538 17193 sgd_solver.cpp:106] Iteration 98000, lr = 6.03659e-06 | |
| I1112 11:01:11.959746 17193 solver.cpp:228] Iteration 99000, loss = 0.00556774 | |
| I1112 11:01:11.959866 17193 solver.cpp:244] Train net output #0: loss = 0.00556944 (* 1 = 0.00556944 loss) | |
| I1112 11:01:11.959879 17193 sgd_solver.cpp:106] Iteration 99000, lr = 5.9908e-06 | |
| I1112 11:03:22.230046 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_100000.caffemodel | |
| I1112 11:03:53.647625 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_100000.solverstate | |
| I1112 11:03:54.198310 17193 solver.cpp:337] Iteration 100000, Testing net (#0) | |
| I1112 11:03:54.198361 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 11:03:54.198370 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 11:04:01.403614 17193 solver.cpp:404] Test net output #0: accuracy = 0.998 | |
| I1112 11:04:01.439766 17193 solver.cpp:228] Iteration 100000, loss = 0.0161267 | |
| I1112 11:04:01.439817 17193 solver.cpp:244] Train net output #0: loss = 0.0161283 (* 1 = 0.0161283 loss) | |
| I1112 11:04:01.439831 17193 sgd_solver.cpp:106] Iteration 100000, lr = 5.94581e-06 | |
| I1112 11:06:11.853008 17193 solver.cpp:228] Iteration 101000, loss = 0.00543628 | |
| I1112 11:06:11.853147 17193 solver.cpp:244] Train net output #0: loss = 0.00543799 (* 1 = 0.00543799 loss) | |
| I1112 11:06:11.853159 17193 sgd_solver.cpp:106] Iteration 101000, lr = 5.90161e-06 | |
| I1112 11:08:22.258002 17193 solver.cpp:228] Iteration 102000, loss = 0.0145269 | |
| I1112 11:08:22.258123 17193 solver.cpp:244] Train net output #0: loss = 0.0145285 (* 1 = 0.0145285 loss) | |
| I1112 11:08:22.258137 17193 sgd_solver.cpp:106] Iteration 102000, lr = 5.85816e-06 | |
| I1112 11:10:32.667886 17193 solver.cpp:228] Iteration 103000, loss = 0.00577882 | |
| I1112 11:10:32.668035 17193 solver.cpp:244] Train net output #0: loss = 0.00578045 (* 1 = 0.00578045 loss) | |
| I1112 11:10:32.668051 17193 sgd_solver.cpp:106] Iteration 103000, lr = 5.81546e-06 | |
| I1112 11:12:43.077463 17193 solver.cpp:228] Iteration 104000, loss = 0.00145848 | |
| I1112 11:12:43.077584 17193 solver.cpp:244] Train net output #0: loss = 0.00146013 (* 1 = 0.00146013 loss) | |
| I1112 11:12:43.077597 17193 sgd_solver.cpp:106] Iteration 104000, lr = 5.77347e-06 | |
| I1112 11:14:53.343333 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_105000.caffemodel | |
| I1112 11:15:12.092910 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_105000.solverstate | |
| I1112 11:15:12.839843 17193 solver.cpp:337] Iteration 105000, Testing net (#0) | |
| I1112 11:15:12.839897 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 11:15:12.839911 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 11:15:20.050802 17193 solver.cpp:404] Test net output #0: accuracy = 0.99744 | |
| I1112 11:15:20.088654 17193 solver.cpp:228] Iteration 105000, loss = 0.00872987 | |
| I1112 11:15:20.088734 17193 solver.cpp:244] Train net output #0: loss = 0.00873159 (* 1 = 0.00873159 loss) | |
| I1112 11:15:20.088755 17193 sgd_solver.cpp:106] Iteration 105000, lr = 5.73218e-06 | |
| I1112 11:17:30.539573 17193 solver.cpp:228] Iteration 106000, loss = 0.01071 | |
| I1112 11:17:30.539649 17193 solver.cpp:244] Train net output #0: loss = 0.0107117 (* 1 = 0.0107117 loss) | |
| I1112 11:17:30.539661 17193 sgd_solver.cpp:106] Iteration 106000, lr = 5.69158e-06 | |
| I1112 11:19:40.968560 17193 solver.cpp:228] Iteration 107000, loss = 0.0138835 | |
| I1112 11:19:40.968700 17193 solver.cpp:244] Train net output #0: loss = 0.0138853 (* 1 = 0.0138853 loss) | |
| I1112 11:19:40.968713 17193 sgd_solver.cpp:106] Iteration 107000, lr = 5.65164e-06 | |
| I1112 11:21:51.371809 17193 solver.cpp:228] Iteration 108000, loss = 0.0127865 | |
| I1112 11:21:51.371949 17193 solver.cpp:244] Train net output #0: loss = 0.0127882 (* 1 = 0.0127882 loss) | |
| I1112 11:21:51.371961 17193 sgd_solver.cpp:106] Iteration 108000, lr = 5.61235e-06 | |
| I1112 11:24:01.776815 17193 solver.cpp:228] Iteration 109000, loss = 0.00385717 | |
| I1112 11:24:01.776906 17193 solver.cpp:244] Train net output #0: loss = 0.00385894 (* 1 = 0.00385894 loss) | |
| I1112 11:24:01.776919 17193 sgd_solver.cpp:106] Iteration 109000, lr = 5.57369e-06 | |
| I1112 11:26:12.048195 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_110000.caffemodel | |
| I1112 11:26:35.827205 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_110000.solverstate | |
| I1112 11:26:36.474148 17193 solver.cpp:337] Iteration 110000, Testing net (#0) | |
| I1112 11:26:36.474189 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 11:26:36.474210 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 11:26:43.675721 17193 solver.cpp:404] Test net output #0: accuracy = 0.99776 | |
| I1112 11:26:43.715004 17193 solver.cpp:228] Iteration 110000, loss = 0.0056683 | |
| I1112 11:26:43.715055 17193 solver.cpp:244] Train net output #0: loss = 0.00567008 (* 1 = 0.00567008 loss) | |
| I1112 11:26:43.715067 17193 sgd_solver.cpp:106] Iteration 110000, lr = 5.53564e-06 | |
| I1112 11:28:54.115092 17193 solver.cpp:228] Iteration 111000, loss = 0.00765356 | |
| I1112 11:28:54.115165 17193 solver.cpp:244] Train net output #0: loss = 0.00765531 (* 1 = 0.00765531 loss) | |
| I1112 11:28:54.115177 17193 sgd_solver.cpp:106] Iteration 111000, lr = 5.4982e-06 | |
| I1112 11:31:04.588783 17193 solver.cpp:228] Iteration 112000, loss = 0.0160112 | |
| I1112 11:31:04.588873 17193 solver.cpp:244] Train net output #0: loss = 0.016013 (* 1 = 0.016013 loss) | |
| I1112 11:31:04.588886 17193 sgd_solver.cpp:106] Iteration 112000, lr = 5.46134e-06 | |
| I1112 11:33:14.996325 17193 solver.cpp:228] Iteration 113000, loss = 0.0122499 | |
| I1112 11:33:14.996443 17193 solver.cpp:244] Train net output #0: loss = 0.0122516 (* 1 = 0.0122516 loss) | |
| I1112 11:33:14.996455 17193 sgd_solver.cpp:106] Iteration 113000, lr = 5.42506e-06 | |
| I1112 11:35:25.420202 17193 solver.cpp:228] Iteration 114000, loss = 0.00470987 | |
| I1112 11:35:25.420377 17193 solver.cpp:244] Train net output #0: loss = 0.00471159 (* 1 = 0.00471159 loss) | |
| I1112 11:35:25.420390 17193 sgd_solver.cpp:106] Iteration 114000, lr = 5.38933e-06 | |
| I1112 11:37:35.704777 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_115000.caffemodel | |
| I1112 11:38:02.572888 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_115000.solverstate | |
| I1112 11:38:02.711416 17193 solver.cpp:337] Iteration 115000, Testing net (#0) | |
| I1112 11:38:02.711465 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 11:38:02.711483 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 11:38:09.909050 17193 solver.cpp:404] Test net output #0: accuracy = 0.9976 | |
| I1112 11:38:09.944818 17193 solver.cpp:228] Iteration 115000, loss = 0.00276966 | |
| I1112 11:38:09.944871 17193 solver.cpp:244] Train net output #0: loss = 0.00277137 (* 1 = 0.00277137 loss) | |
| I1112 11:38:09.944887 17193 sgd_solver.cpp:106] Iteration 115000, lr = 5.35414e-06 | |
| I1112 11:40:20.345904 17193 solver.cpp:228] Iteration 116000, loss = 0.0170964 | |
| I1112 11:40:20.345999 17193 solver.cpp:244] Train net output #0: loss = 0.0170981 (* 1 = 0.0170981 loss) | |
| I1112 11:40:20.346009 17193 sgd_solver.cpp:106] Iteration 116000, lr = 5.31949e-06 | |
| I1112 11:42:30.741521 17193 solver.cpp:228] Iteration 117000, loss = 0.00253313 | |
| I1112 11:42:30.741606 17193 solver.cpp:244] Train net output #0: loss = 0.00253483 (* 1 = 0.00253483 loss) | |
| I1112 11:42:30.741618 17193 sgd_solver.cpp:106] Iteration 117000, lr = 5.28536e-06 | |
| I1112 11:44:41.134482 17193 solver.cpp:228] Iteration 118000, loss = 0.00764648 | |
| I1112 11:44:41.134574 17193 solver.cpp:244] Train net output #0: loss = 0.00764813 (* 1 = 0.00764813 loss) | |
| I1112 11:44:41.134588 17193 sgd_solver.cpp:106] Iteration 118000, lr = 5.25173e-06 | |
| I1112 11:46:51.535643 17193 solver.cpp:228] Iteration 119000, loss = 0.00309567 | |
| I1112 11:46:51.535729 17193 solver.cpp:244] Train net output #0: loss = 0.00309737 (* 1 = 0.00309737 loss) | |
| I1112 11:46:51.535740 17193 sgd_solver.cpp:106] Iteration 119000, lr = 5.2186e-06 | |
| I1112 11:49:01.816495 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_120000.caffemodel | |
| I1112 11:49:19.828313 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_120000.solverstate | |
| I1112 11:49:19.966259 17193 solver.cpp:337] Iteration 120000, Testing net (#0) | |
| I1112 11:49:19.966310 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 11:49:19.966320 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 11:49:27.169723 17193 solver.cpp:404] Test net output #0: accuracy = 0.9976 | |
| I1112 11:49:27.206605 17193 solver.cpp:228] Iteration 120000, loss = 0.0134996 | |
| I1112 11:49:27.206658 17193 solver.cpp:244] Train net output #0: loss = 0.0135014 (* 1 = 0.0135014 loss) | |
| I1112 11:49:27.206673 17193 sgd_solver.cpp:106] Iteration 120000, lr = 5.18595e-06 | |
| I1112 11:51:37.628993 17193 solver.cpp:228] Iteration 121000, loss = 0.00552269 | |
| I1112 11:51:37.629118 17193 solver.cpp:244] Train net output #0: loss = 0.00552434 (* 1 = 0.00552434 loss) | |
| I1112 11:51:37.629133 17193 sgd_solver.cpp:106] Iteration 121000, lr = 5.15377e-06 | |
| I1112 11:53:48.051872 17193 solver.cpp:228] Iteration 122000, loss = 0.0153007 | |
| I1112 11:53:48.051959 17193 solver.cpp:244] Train net output #0: loss = 0.0153025 (* 1 = 0.0153025 loss) | |
| I1112 11:53:48.051972 17193 sgd_solver.cpp:106] Iteration 122000, lr = 5.12206e-06 | |
| I1112 11:55:58.467332 17193 solver.cpp:228] Iteration 123000, loss = 0.00218341 | |
| I1112 11:55:58.467456 17193 solver.cpp:244] Train net output #0: loss = 0.00218512 (* 1 = 0.00218512 loss) | |
| I1112 11:55:58.467484 17193 sgd_solver.cpp:106] Iteration 123000, lr = 5.09079e-06 | |
| I1112 11:58:08.880429 17193 solver.cpp:228] Iteration 124000, loss = 0.0047274 | |
| I1112 11:58:08.880516 17193 solver.cpp:244] Train net output #0: loss = 0.00472914 (* 1 = 0.00472914 loss) | |
| I1112 11:58:08.880528 17193 sgd_solver.cpp:106] Iteration 124000, lr = 5.05997e-06 | |
| I1112 12:00:19.155210 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_125000.caffemodel | |
| I1112 12:00:46.252993 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_125000.solverstate | |
| I1112 12:00:46.390389 17193 solver.cpp:337] Iteration 125000, Testing net (#0) | |
| I1112 12:00:46.390439 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:00:46.390450 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:00:53.596827 17193 solver.cpp:404] Test net output #0: accuracy = 0.99768 | |
| I1112 12:00:53.632899 17193 solver.cpp:228] Iteration 125000, loss = 0.00964669 | |
| I1112 12:00:53.632949 17193 solver.cpp:244] Train net output #0: loss = 0.00964844 (* 1 = 0.00964844 loss) | |
| I1112 12:00:53.632964 17193 sgd_solver.cpp:106] Iteration 125000, lr = 5.02958e-06 | |
| I1112 12:03:04.042045 17193 solver.cpp:228] Iteration 126000, loss = 0.00579223 | |
| I1112 12:03:04.042120 17193 solver.cpp:244] Train net output #0: loss = 0.00579399 (* 1 = 0.00579399 loss) | |
| I1112 12:03:04.042135 17193 sgd_solver.cpp:106] Iteration 126000, lr = 4.99962e-06 | |
| I1112 12:05:14.444810 17193 solver.cpp:228] Iteration 127000, loss = 0.0195434 | |
| I1112 12:05:14.444916 17193 solver.cpp:244] Train net output #0: loss = 0.0195452 (* 1 = 0.0195452 loss) | |
| I1112 12:05:14.444941 17193 sgd_solver.cpp:106] Iteration 127000, lr = 4.97006e-06 | |
| I1112 12:07:24.854748 17193 solver.cpp:228] Iteration 128000, loss = 0.0196261 | |
| I1112 12:07:24.854856 17193 solver.cpp:244] Train net output #0: loss = 0.0196279 (* 1 = 0.0196279 loss) | |
| I1112 12:07:24.854879 17193 sgd_solver.cpp:106] Iteration 128000, lr = 4.94091e-06 | |
| I1112 12:09:35.286113 17193 solver.cpp:228] Iteration 129000, loss = 0.00688472 | |
| I1112 12:09:35.286216 17193 solver.cpp:244] Train net output #0: loss = 0.00688655 (* 1 = 0.00688655 loss) | |
| I1112 12:09:35.286242 17193 sgd_solver.cpp:106] Iteration 129000, lr = 4.91216e-06 | |
| I1112 12:11:45.542387 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_130000.caffemodel | |
| I1112 12:12:08.724977 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_130000.solverstate | |
| I1112 12:12:08.865871 17193 solver.cpp:337] Iteration 130000, Testing net (#0) | |
| I1112 12:12:08.865924 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:12:08.865933 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:12:16.067539 17193 solver.cpp:404] Test net output #0: accuracy = 0.99808 | |
| I1112 12:12:16.103832 17193 solver.cpp:228] Iteration 130000, loss = 0.0189293 | |
| I1112 12:12:16.103886 17193 solver.cpp:244] Train net output #0: loss = 0.0189312 (* 1 = 0.0189312 loss) | |
| I1112 12:12:16.103901 17193 sgd_solver.cpp:106] Iteration 130000, lr = 4.8838e-06 | |
| I1112 12:14:26.488951 17193 solver.cpp:228] Iteration 131000, loss = 0.0022104 | |
| I1112 12:14:26.489056 17193 solver.cpp:244] Train net output #0: loss = 0.0022122 (* 1 = 0.0022122 loss) | |
| I1112 12:14:26.489080 17193 sgd_solver.cpp:106] Iteration 131000, lr = 4.85581e-06 | |
| I1112 12:16:36.878696 17193 solver.cpp:228] Iteration 132000, loss = 0.00378329 | |
| I1112 12:16:36.878803 17193 solver.cpp:244] Train net output #0: loss = 0.00378512 (* 1 = 0.00378512 loss) | |
| I1112 12:16:36.878826 17193 sgd_solver.cpp:106] Iteration 132000, lr = 4.82819e-06 | |
| I1112 12:18:47.278796 17193 solver.cpp:228] Iteration 133000, loss = 0.00410104 | |
| I1112 12:18:47.278985 17193 solver.cpp:244] Train net output #0: loss = 0.0041029 (* 1 = 0.0041029 loss) | |
| I1112 12:18:47.279021 17193 sgd_solver.cpp:106] Iteration 133000, lr = 4.80094e-06 | |
| I1112 12:20:57.698715 17193 solver.cpp:228] Iteration 134000, loss = 0.00228239 | |
| I1112 12:20:57.698819 17193 solver.cpp:244] Train net output #0: loss = 0.00228427 (* 1 = 0.00228427 loss) | |
| I1112 12:20:57.698844 17193 sgd_solver.cpp:106] Iteration 134000, lr = 4.77405e-06 | |
| I1112 12:23:07.992764 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_135000.caffemodel | |
| I1112 12:23:36.272923 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_135000.solverstate | |
| I1112 12:23:36.412394 17193 solver.cpp:337] Iteration 135000, Testing net (#0) | |
| I1112 12:23:36.412444 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:23:36.412453 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:23:43.606752 17193 solver.cpp:404] Test net output #0: accuracy = 0.99824 | |
| I1112 12:23:43.643174 17193 solver.cpp:228] Iteration 135000, loss = 0.0034159 | |
| I1112 12:23:43.643225 17193 solver.cpp:244] Train net output #0: loss = 0.00341779 (* 1 = 0.00341779 loss) | |
| I1112 12:23:43.643239 17193 sgd_solver.cpp:106] Iteration 135000, lr = 4.7475e-06 | |
| I1112 12:25:54.040248 17193 solver.cpp:228] Iteration 136000, loss = 0.0120066 | |
| I1112 12:25:54.040343 17193 solver.cpp:244] Train net output #0: loss = 0.0120085 (* 1 = 0.0120085 loss) | |
| I1112 12:25:54.040355 17193 sgd_solver.cpp:106] Iteration 136000, lr = 4.7213e-06 | |
| I1112 12:28:04.452798 17193 solver.cpp:228] Iteration 137000, loss = 0.00464187 | |
| I1112 12:28:04.452875 17193 solver.cpp:244] Train net output #0: loss = 0.00464369 (* 1 = 0.00464369 loss) | |
| I1112 12:28:04.452890 17193 sgd_solver.cpp:106] Iteration 137000, lr = 4.69543e-06 | |
| I1112 12:30:14.839046 17193 solver.cpp:228] Iteration 138000, loss = 0.00190926 | |
| I1112 12:30:14.839123 17193 solver.cpp:244] Train net output #0: loss = 0.00191105 (* 1 = 0.00191105 loss) | |
| I1112 12:30:14.839135 17193 sgd_solver.cpp:106] Iteration 138000, lr = 4.66989e-06 | |
| I1112 12:32:25.248997 17193 solver.cpp:228] Iteration 139000, loss = 0.0552999 | |
| I1112 12:32:25.249105 17193 solver.cpp:244] Train net output #0: loss = 0.0553017 (* 1 = 0.0553017 loss) | |
| I1112 12:32:25.249130 17193 sgd_solver.cpp:106] Iteration 139000, lr = 4.64467e-06 | |
| I1112 12:34:35.532831 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_140000.caffemodel | |
| I1112 12:35:09.020220 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_140000.solverstate | |
| I1112 12:35:10.204108 17193 solver.cpp:337] Iteration 140000, Testing net (#0) | |
| I1112 12:35:10.204159 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:35:10.204166 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:35:17.407395 17193 solver.cpp:404] Test net output #0: accuracy = 0.998 | |
| I1112 12:35:17.442706 17193 solver.cpp:228] Iteration 140000, loss = 0.0363352 | |
| I1112 12:35:17.442750 17193 solver.cpp:244] Train net output #0: loss = 0.036337 (* 1 = 0.036337 loss) | |
| I1112 12:35:17.442765 17193 sgd_solver.cpp:106] Iteration 140000, lr = 4.61976e-06 | |
| I1112 12:37:27.845494 17193 solver.cpp:228] Iteration 141000, loss = 0.0526587 | |
| I1112 12:37:27.845640 17193 solver.cpp:244] Train net output #0: loss = 0.0526605 (* 1 = 0.0526605 loss) | |
| I1112 12:37:27.845655 17193 sgd_solver.cpp:106] Iteration 141000, lr = 4.59517e-06 | |
| I1112 12:39:38.265979 17193 solver.cpp:228] Iteration 142000, loss = 0.00281541 | |
| I1112 12:39:38.266062 17193 solver.cpp:244] Train net output #0: loss = 0.00281726 (* 1 = 0.00281726 loss) | |
| I1112 12:39:38.266073 17193 sgd_solver.cpp:106] Iteration 142000, lr = 4.57088e-06 | |
| I1112 12:41:48.681813 17193 solver.cpp:228] Iteration 143000, loss = 0.00215914 | |
| I1112 12:41:48.681910 17193 solver.cpp:244] Train net output #0: loss = 0.00216099 (* 1 = 0.00216099 loss) | |
| I1112 12:41:48.681921 17193 sgd_solver.cpp:106] Iteration 143000, lr = 4.54689e-06 | |
| I1112 12:43:59.115756 17193 solver.cpp:228] Iteration 144000, loss = 0.00389099 | |
| I1112 12:43:59.115820 17193 solver.cpp:244] Train net output #0: loss = 0.00389287 (* 1 = 0.00389287 loss) | |
| I1112 12:43:59.115831 17193 sgd_solver.cpp:106] Iteration 144000, lr = 4.52318e-06 | |
| I1112 12:46:09.419272 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_145000.caffemodel | |
| I1112 12:46:33.935945 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_145000.solverstate | |
| I1112 12:46:34.075363 17193 solver.cpp:337] Iteration 145000, Testing net (#0) | |
| I1112 12:46:34.075412 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:46:34.075420 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:46:41.277225 17193 solver.cpp:404] Test net output #0: accuracy = 0.99776 | |
| I1112 12:46:41.313522 17193 solver.cpp:228] Iteration 145000, loss = 0.00208943 | |
| I1112 12:46:41.313571 17193 solver.cpp:244] Train net output #0: loss = 0.00209132 (* 1 = 0.00209132 loss) | |
| I1112 12:46:41.313583 17193 sgd_solver.cpp:106] Iteration 145000, lr = 4.49977e-06 | |
| I1112 12:48:51.714341 17193 solver.cpp:228] Iteration 146000, loss = 0.00641618 | |
| I1112 12:48:51.714450 17193 solver.cpp:244] Train net output #0: loss = 0.00641806 (* 1 = 0.00641806 loss) | |
| I1112 12:48:51.714473 17193 sgd_solver.cpp:106] Iteration 146000, lr = 4.47664e-06 | |
| I1112 12:51:02.116539 17193 solver.cpp:228] Iteration 147000, loss = 0.00921038 | |
| I1112 12:51:02.116726 17193 solver.cpp:244] Train net output #0: loss = 0.00921227 (* 1 = 0.00921227 loss) | |
| I1112 12:51:02.116742 17193 sgd_solver.cpp:106] Iteration 147000, lr = 4.45378e-06 | |
| I1112 12:53:12.526419 17193 solver.cpp:228] Iteration 148000, loss = 0.00627513 | |
| I1112 12:53:12.526491 17193 solver.cpp:244] Train net output #0: loss = 0.006277 (* 1 = 0.006277 loss) | |
| I1112 12:53:12.526504 17193 sgd_solver.cpp:106] Iteration 148000, lr = 4.43119e-06 | |
| I1112 12:55:22.932384 17193 solver.cpp:228] Iteration 149000, loss = 0.0643277 | |
| I1112 12:55:22.932458 17193 solver.cpp:244] Train net output #0: loss = 0.0643296 (* 1 = 0.0643296 loss) | |
| I1112 12:55:22.932471 17193 sgd_solver.cpp:106] Iteration 149000, lr = 4.40887e-06 | |
| I1112 12:57:33.228235 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_150000.caffemodel | |
| I1112 12:57:57.830957 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_150000.solverstate | |
| I1112 12:57:57.971102 17193 solver.cpp:337] Iteration 150000, Testing net (#0) | |
| I1112 12:57:57.971141 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 12:57:57.971161 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 12:58:05.180610 17193 solver.cpp:404] Test net output #0: accuracy = 0.99856 | |
| I1112 12:58:05.216403 17193 solver.cpp:228] Iteration 150000, loss = 0.00284142 | |
| I1112 12:58:05.216451 17193 solver.cpp:244] Train net output #0: loss = 0.00284326 (* 1 = 0.00284326 loss) | |
| I1112 12:58:05.216465 17193 sgd_solver.cpp:106] Iteration 150000, lr = 4.3868e-06 | |
| I1112 13:00:15.585351 17193 solver.cpp:228] Iteration 151000, loss = 0.00477281 | |
| I1112 13:00:15.585471 17193 solver.cpp:244] Train net output #0: loss = 0.00477465 (* 1 = 0.00477465 loss) | |
| I1112 13:00:15.585485 17193 sgd_solver.cpp:106] Iteration 151000, lr = 4.365e-06 | |
| I1112 13:02:25.972086 17193 solver.cpp:228] Iteration 152000, loss = 0.0510125 | |
| I1112 13:02:25.972205 17193 solver.cpp:244] Train net output #0: loss = 0.0510144 (* 1 = 0.0510144 loss) | |
| I1112 13:02:25.972218 17193 sgd_solver.cpp:106] Iteration 152000, lr = 4.34344e-06 | |
| I1112 13:04:36.369133 17193 solver.cpp:228] Iteration 153000, loss = 0.00210276 | |
| I1112 13:04:36.369204 17193 solver.cpp:244] Train net output #0: loss = 0.0021046 (* 1 = 0.0021046 loss) | |
| I1112 13:04:36.369217 17193 sgd_solver.cpp:106] Iteration 153000, lr = 4.32213e-06 | |
| I1112 13:06:46.779214 17193 solver.cpp:228] Iteration 154000, loss = 0.00535677 | |
| I1112 13:06:46.779333 17193 solver.cpp:244] Train net output #0: loss = 0.00535861 (* 1 = 0.00535861 loss) | |
| I1112 13:06:46.779345 17193 sgd_solver.cpp:106] Iteration 154000, lr = 4.30107e-06 | |
| I1112 13:08:57.067198 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_155000.caffemodel | |
| I1112 13:09:23.219465 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_155000.solverstate | |
| I1112 13:09:23.356142 17193 solver.cpp:337] Iteration 155000, Testing net (#0) | |
| I1112 13:09:23.356191 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 13:09:23.356200 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 13:09:30.563210 17193 solver.cpp:404] Test net output #0: accuracy = 0.99848 | |
| I1112 13:09:30.599601 17193 solver.cpp:228] Iteration 155000, loss = 0.00396648 | |
| I1112 13:09:30.599650 17193 solver.cpp:244] Train net output #0: loss = 0.00396833 (* 1 = 0.00396833 loss) | |
| I1112 13:09:30.599665 17193 sgd_solver.cpp:106] Iteration 155000, lr = 4.28024e-06 | |
| I1112 13:11:40.990519 17193 solver.cpp:228] Iteration 156000, loss = 0.0100048 | |
| I1112 13:11:40.990651 17193 solver.cpp:244] Train net output #0: loss = 0.0100067 (* 1 = 0.0100067 loss) | |
| I1112 13:11:40.990676 17193 sgd_solver.cpp:106] Iteration 156000, lr = 4.25965e-06 | |
| I1112 13:13:51.405164 17193 solver.cpp:228] Iteration 157000, loss = 0.00986193 | |
| I1112 13:13:51.405242 17193 solver.cpp:244] Train net output #0: loss = 0.00986382 (* 1 = 0.00986382 loss) | |
| I1112 13:13:51.405256 17193 sgd_solver.cpp:106] Iteration 157000, lr = 4.23928e-06 | |
| I1112 13:16:01.810799 17193 solver.cpp:228] Iteration 158000, loss = 0.0223931 | |
| I1112 13:16:01.810885 17193 solver.cpp:244] Train net output #0: loss = 0.0223951 (* 1 = 0.0223951 loss) | |
| I1112 13:16:01.810899 17193 sgd_solver.cpp:106] Iteration 158000, lr = 4.21914e-06 | |
| I1112 13:18:12.210327 17193 solver.cpp:228] Iteration 159000, loss = 0.00839358 | |
| I1112 13:18:12.210413 17193 solver.cpp:244] Train net output #0: loss = 0.00839548 (* 1 = 0.00839548 loss) | |
| I1112 13:18:12.210424 17193 sgd_solver.cpp:106] Iteration 159000, lr = 4.19923e-06 | |
| I1112 13:20:22.471000 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_160000.caffemodel | |
| I1112 13:20:49.264866 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_160000.solverstate | |
| I1112 13:20:49.403817 17193 solver.cpp:337] Iteration 160000, Testing net (#0) | |
| I1112 13:20:49.403867 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 13:20:49.403877 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 13:20:56.614260 17193 solver.cpp:404] Test net output #0: accuracy = 0.99792 | |
| I1112 13:20:56.650038 17193 solver.cpp:228] Iteration 160000, loss = 0.00453755 | |
| I1112 13:20:56.650092 17193 solver.cpp:244] Train net output #0: loss = 0.00453946 (* 1 = 0.00453946 loss) | |
| I1112 13:20:56.650108 17193 sgd_solver.cpp:106] Iteration 160000, lr = 4.17953e-06 | |
| I1112 13:23:07.070565 17193 solver.cpp:228] Iteration 161000, loss = 0.0266744 | |
| I1112 13:23:07.070658 17193 solver.cpp:244] Train net output #0: loss = 0.0266763 (* 1 = 0.0266763 loss) | |
| I1112 13:23:07.070670 17193 sgd_solver.cpp:106] Iteration 161000, lr = 4.16004e-06 | |
| I1112 13:25:17.459127 17193 solver.cpp:228] Iteration 162000, loss = 0.00684146 | |
| I1112 13:25:17.459218 17193 solver.cpp:244] Train net output #0: loss = 0.00684339 (* 1 = 0.00684339 loss) | |
| I1112 13:25:17.459228 17193 sgd_solver.cpp:106] Iteration 162000, lr = 4.14077e-06 | |
| I1112 13:27:27.866778 17193 solver.cpp:228] Iteration 163000, loss = 0.0066272 | |
| I1112 13:27:27.866869 17193 solver.cpp:244] Train net output #0: loss = 0.0066291 (* 1 = 0.0066291 loss) | |
| I1112 13:27:27.866881 17193 sgd_solver.cpp:106] Iteration 163000, lr = 4.1217e-06 | |
| I1112 13:29:38.279705 17193 solver.cpp:228] Iteration 164000, loss = 0.0144103 | |
| I1112 13:29:38.279796 17193 solver.cpp:244] Train net output #0: loss = 0.0144122 (* 1 = 0.0144122 loss) | |
| I1112 13:29:38.279808 17193 sgd_solver.cpp:106] Iteration 164000, lr = 4.10284e-06 | |
| I1112 13:31:48.554280 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_165000.caffemodel | |
| I1112 13:32:22.073691 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_165000.solverstate | |
| I1112 13:32:22.213836 17193 solver.cpp:337] Iteration 165000, Testing net (#0) | |
| I1112 13:32:22.213887 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 13:32:22.213896 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 13:32:29.415191 17193 solver.cpp:404] Test net output #0: accuracy = 0.998 | |
| I1112 13:32:29.450726 17193 solver.cpp:228] Iteration 165000, loss = 0.00275739 | |
| I1112 13:32:29.450779 17193 solver.cpp:244] Train net output #0: loss = 0.00275929 (* 1 = 0.00275929 loss) | |
| I1112 13:32:29.450794 17193 sgd_solver.cpp:106] Iteration 165000, lr = 4.08418e-06 | |
| I1112 13:34:39.849742 17193 solver.cpp:228] Iteration 166000, loss = 0.0180023 | |
| I1112 13:34:39.849862 17193 solver.cpp:244] Train net output #0: loss = 0.0180042 (* 1 = 0.0180042 loss) | |
| I1112 13:34:39.849876 17193 sgd_solver.cpp:106] Iteration 166000, lr = 4.06571e-06 | |
| I1112 13:36:50.266611 17193 solver.cpp:228] Iteration 167000, loss = 0.00940369 | |
| I1112 13:36:50.266705 17193 solver.cpp:244] Train net output #0: loss = 0.00940567 (* 1 = 0.00940567 loss) | |
| I1112 13:36:50.266721 17193 sgd_solver.cpp:106] Iteration 167000, lr = 4.04744e-06 | |
| I1112 13:39:00.675823 17193 solver.cpp:228] Iteration 168000, loss = 0.00118247 | |
| I1112 13:39:00.675906 17193 solver.cpp:244] Train net output #0: loss = 0.00118441 (* 1 = 0.00118441 loss) | |
| I1112 13:39:00.675917 17193 sgd_solver.cpp:106] Iteration 168000, lr = 4.02936e-06 | |
| I1112 13:41:11.082743 17193 solver.cpp:228] Iteration 169000, loss = 0.00113313 | |
| I1112 13:41:11.082814 17193 solver.cpp:244] Train net output #0: loss = 0.00113505 (* 1 = 0.00113505 loss) | |
| I1112 13:41:11.082826 17193 sgd_solver.cpp:106] Iteration 169000, lr = 4.01146e-06 | |
| I1112 13:43:21.367377 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_170000.caffemodel | |
| I1112 13:43:44.857727 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_170000.solverstate | |
| I1112 13:43:44.996640 17193 solver.cpp:337] Iteration 170000, Testing net (#0) | |
| I1112 13:43:44.996691 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 13:43:44.996700 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 13:43:52.188726 17193 solver.cpp:404] Test net output #0: accuracy = 0.99816 | |
| I1112 13:43:52.225473 17193 solver.cpp:228] Iteration 170000, loss = 0.00484233 | |
| I1112 13:43:52.225527 17193 solver.cpp:244] Train net output #0: loss = 0.00484425 (* 1 = 0.00484425 loss) | |
| I1112 13:43:52.225543 17193 sgd_solver.cpp:106] Iteration 170000, lr = 3.99375e-06 | |
| I1112 13:46:02.637689 17193 solver.cpp:228] Iteration 171000, loss = 0.00115312 | |
| I1112 13:46:02.637819 17193 solver.cpp:244] Train net output #0: loss = 0.00115499 (* 1 = 0.00115499 loss) | |
| I1112 13:46:02.637831 17193 sgd_solver.cpp:106] Iteration 171000, lr = 3.97622e-06 | |
| I1112 13:48:13.054709 17193 solver.cpp:228] Iteration 172000, loss = 0.00362315 | |
| I1112 13:48:13.054800 17193 solver.cpp:244] Train net output #0: loss = 0.00362504 (* 1 = 0.00362504 loss) | |
| I1112 13:48:13.054814 17193 sgd_solver.cpp:106] Iteration 172000, lr = 3.95887e-06 | |
| I1112 13:50:23.471549 17193 solver.cpp:228] Iteration 173000, loss = 0.00207071 | |
| I1112 13:50:23.471642 17193 solver.cpp:244] Train net output #0: loss = 0.00207262 (* 1 = 0.00207262 loss) | |
| I1112 13:50:23.471658 17193 sgd_solver.cpp:106] Iteration 173000, lr = 3.9417e-06 | |
| I1112 13:52:33.881455 17193 solver.cpp:228] Iteration 174000, loss = 0.00317128 | |
| I1112 13:52:33.881547 17193 solver.cpp:244] Train net output #0: loss = 0.0031732 (* 1 = 0.0031732 loss) | |
| I1112 13:52:33.881559 17193 sgd_solver.cpp:106] Iteration 174000, lr = 3.9247e-06 | |
| I1112 13:54:44.145556 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_175000.caffemodel | |
| I1112 13:55:18.169812 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_175000.solverstate | |
| I1112 13:55:18.308523 17193 solver.cpp:337] Iteration 175000, Testing net (#0) | |
| I1112 13:55:18.308573 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 13:55:18.308581 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 13:55:25.506369 17193 solver.cpp:404] Test net output #0: accuracy = 0.9984 | |
| I1112 13:55:25.541895 17193 solver.cpp:228] Iteration 175000, loss = 0.019802 | |
| I1112 13:55:25.541941 17193 solver.cpp:244] Train net output #0: loss = 0.019804 (* 1 = 0.019804 loss) | |
| I1112 13:55:25.541956 17193 sgd_solver.cpp:106] Iteration 175000, lr = 3.90787e-06 | |
| I1112 13:57:35.932982 17193 solver.cpp:228] Iteration 176000, loss = 0.0364714 | |
| I1112 13:57:35.933084 17193 solver.cpp:244] Train net output #0: loss = 0.0364734 (* 1 = 0.0364734 loss) | |
| I1112 13:57:35.933107 17193 sgd_solver.cpp:106] Iteration 176000, lr = 3.8912e-06 | |
| I1112 13:59:46.346225 17193 solver.cpp:228] Iteration 177000, loss = 0.0167498 | |
| I1112 13:59:46.346341 17193 solver.cpp:244] Train net output #0: loss = 0.0167518 (* 1 = 0.0167518 loss) | |
| I1112 13:59:46.346354 17193 sgd_solver.cpp:106] Iteration 177000, lr = 3.8747e-06 | |
| I1112 14:01:56.745548 17193 solver.cpp:228] Iteration 178000, loss = 0.0024755 | |
| I1112 14:01:56.745656 17193 solver.cpp:244] Train net output #0: loss = 0.00247749 (* 1 = 0.00247749 loss) | |
| I1112 14:01:56.745679 17193 sgd_solver.cpp:106] Iteration 178000, lr = 3.85837e-06 | |
| I1112 14:04:07.154556 17193 solver.cpp:228] Iteration 179000, loss = 0.0103652 | |
| I1112 14:04:07.154656 17193 solver.cpp:244] Train net output #0: loss = 0.0103672 (* 1 = 0.0103672 loss) | |
| I1112 14:04:07.154680 17193 sgd_solver.cpp:106] Iteration 179000, lr = 3.84219e-06 | |
| I1112 14:06:17.423765 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_180000.caffemodel | |
| I1112 14:06:44.626078 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_180000.solverstate | |
| I1112 14:06:45.114624 17193 solver.cpp:337] Iteration 180000, Testing net (#0) | |
| I1112 14:06:45.114677 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 14:06:45.114686 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 14:06:52.326354 17193 solver.cpp:404] Test net output #0: accuracy = 0.99864 | |
| I1112 14:06:52.362536 17193 solver.cpp:228] Iteration 180000, loss = 0.00127268 | |
| I1112 14:06:52.362591 17193 solver.cpp:244] Train net output #0: loss = 0.0012747 (* 1 = 0.0012747 loss) | |
| I1112 14:06:52.362607 17193 sgd_solver.cpp:106] Iteration 180000, lr = 3.82617e-06 | |
| I1112 14:09:02.778605 17193 solver.cpp:228] Iteration 181000, loss = 0.00491031 | |
| I1112 14:09:02.778676 17193 solver.cpp:244] Train net output #0: loss = 0.00491233 (* 1 = 0.00491233 loss) | |
| I1112 14:09:02.778690 17193 sgd_solver.cpp:106] Iteration 181000, lr = 3.8103e-06 | |
| I1112 14:11:13.199157 17193 solver.cpp:228] Iteration 182000, loss = 0.00229318 | |
| I1112 14:11:13.199239 17193 solver.cpp:244] Train net output #0: loss = 0.00229517 (* 1 = 0.00229517 loss) | |
| I1112 14:11:13.199254 17193 sgd_solver.cpp:106] Iteration 182000, lr = 3.79459e-06 | |
| I1112 14:13:23.609714 17193 solver.cpp:228] Iteration 183000, loss = 0.00537014 | |
| I1112 14:13:23.609818 17193 solver.cpp:244] Train net output #0: loss = 0.00537213 (* 1 = 0.00537213 loss) | |
| I1112 14:13:23.609843 17193 sgd_solver.cpp:106] Iteration 183000, lr = 3.77903e-06 | |
| I1112 14:15:34.022562 17193 solver.cpp:228] Iteration 184000, loss = 0.00498535 | |
| I1112 14:15:34.022645 17193 solver.cpp:244] Train net output #0: loss = 0.00498734 (* 1 = 0.00498734 loss) | |
| I1112 14:15:34.022657 17193 sgd_solver.cpp:106] Iteration 184000, lr = 3.76362e-06 | |
| I1112 14:17:44.318693 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_185000.caffemodel | |
| I1112 14:18:18.603253 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_185000.solverstate | |
| I1112 14:18:18.742889 17193 solver.cpp:337] Iteration 185000, Testing net (#0) | |
| I1112 14:18:18.742936 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 14:18:18.742944 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 14:18:25.940750 17193 solver.cpp:404] Test net output #0: accuracy = 0.99832 | |
| I1112 14:18:25.978945 17193 solver.cpp:228] Iteration 185000, loss = 0.00899024 | |
| I1112 14:18:25.978996 17193 solver.cpp:244] Train net output #0: loss = 0.00899222 (* 1 = 0.00899222 loss) | |
| I1112 14:18:25.979010 17193 sgd_solver.cpp:106] Iteration 185000, lr = 3.74835e-06 | |
| I1112 14:20:36.377446 17193 solver.cpp:228] Iteration 186000, loss = 0.00132623 | |
| I1112 14:20:36.377537 17193 solver.cpp:244] Train net output #0: loss = 0.00132819 (* 1 = 0.00132819 loss) | |
| I1112 14:20:36.377549 17193 sgd_solver.cpp:106] Iteration 186000, lr = 3.73322e-06 | |
| I1112 14:22:46.792145 17193 solver.cpp:228] Iteration 187000, loss = 0.0123942 | |
| I1112 14:22:46.792232 17193 solver.cpp:244] Train net output #0: loss = 0.0123961 (* 1 = 0.0123961 loss) | |
| I1112 14:22:46.792243 17193 sgd_solver.cpp:106] Iteration 187000, lr = 3.71824e-06 | |
| I1112 14:24:57.221042 17193 solver.cpp:228] Iteration 188000, loss = 0.0152676 | |
| I1112 14:24:57.221150 17193 solver.cpp:244] Train net output #0: loss = 0.0152695 (* 1 = 0.0152695 loss) | |
| I1112 14:24:57.221166 17193 sgd_solver.cpp:106] Iteration 188000, lr = 3.7034e-06 | |
| I1112 14:27:07.632160 17193 solver.cpp:228] Iteration 189000, loss = 0.001701 | |
| I1112 14:27:07.632271 17193 solver.cpp:244] Train net output #0: loss = 0.00170291 (* 1 = 0.00170291 loss) | |
| I1112 14:27:07.632294 17193 sgd_solver.cpp:106] Iteration 189000, lr = 3.68869e-06 | |
| I1112 14:29:17.897347 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_190000.caffemodel | |
| I1112 14:29:49.584887 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_190000.solverstate | |
| I1112 14:29:49.833911 17193 solver.cpp:337] Iteration 190000, Testing net (#0) | |
| I1112 14:29:49.833957 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 14:29:49.833968 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 14:29:57.040351 17193 solver.cpp:404] Test net output #0: accuracy = 0.99856 | |
| I1112 14:29:57.076380 17193 solver.cpp:228] Iteration 190000, loss = 0.00283433 | |
| I1112 14:29:57.076432 17193 solver.cpp:244] Train net output #0: loss = 0.00283622 (* 1 = 0.00283622 loss) | |
| I1112 14:29:57.076447 17193 sgd_solver.cpp:106] Iteration 190000, lr = 3.67412e-06 | |
| I1112 14:32:07.479351 17193 solver.cpp:228] Iteration 191000, loss = 0.00385901 | |
| I1112 14:32:07.479506 17193 solver.cpp:244] Train net output #0: loss = 0.00386094 (* 1 = 0.00386094 loss) | |
| I1112 14:32:07.479521 17193 sgd_solver.cpp:106] Iteration 191000, lr = 3.65969e-06 | |
| I1112 14:34:17.892268 17193 solver.cpp:228] Iteration 192000, loss = 0.0019228 | |
| I1112 14:34:17.892341 17193 solver.cpp:244] Train net output #0: loss = 0.00192478 (* 1 = 0.00192478 loss) | |
| I1112 14:34:17.892355 17193 sgd_solver.cpp:106] Iteration 192000, lr = 3.64538e-06 | |
| I1112 14:36:28.311589 17193 solver.cpp:228] Iteration 193000, loss = 0.00309389 | |
| I1112 14:36:28.311712 17193 solver.cpp:244] Train net output #0: loss = 0.00309589 (* 1 = 0.00309589 loss) | |
| I1112 14:36:28.311728 17193 sgd_solver.cpp:106] Iteration 193000, lr = 3.63121e-06 | |
| I1112 14:38:38.708379 17193 solver.cpp:228] Iteration 194000, loss = 0.00556743 | |
| I1112 14:38:38.708504 17193 solver.cpp:244] Train net output #0: loss = 0.0055694 (* 1 = 0.0055694 loss) | |
| I1112 14:38:38.708519 17193 sgd_solver.cpp:106] Iteration 194000, lr = 3.61716e-06 | |
| I1112 14:40:48.997328 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_195000.caffemodel | |
| I1112 14:41:16.874796 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_195000.solverstate | |
| I1112 14:41:17.405606 17193 solver.cpp:337] Iteration 195000, Testing net (#0) | |
| I1112 14:41:17.405654 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 14:41:17.405738 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 14:41:24.609372 17193 solver.cpp:404] Test net output #0: accuracy = 0.99784 | |
| I1112 14:41:24.645936 17193 solver.cpp:228] Iteration 195000, loss = 0.00878175 | |
| I1112 14:41:24.645985 17193 solver.cpp:244] Train net output #0: loss = 0.00878371 (* 1 = 0.00878371 loss) | |
| I1112 14:41:24.645998 17193 sgd_solver.cpp:106] Iteration 195000, lr = 3.60324e-06 | |
| I1112 14:43:35.080152 17193 solver.cpp:228] Iteration 196000, loss = 0.00639998 | |
| I1112 14:43:35.080296 17193 solver.cpp:244] Train net output #0: loss = 0.00640191 (* 1 = 0.00640191 loss) | |
| I1112 14:43:35.080308 17193 sgd_solver.cpp:106] Iteration 196000, lr = 3.58944e-06 | |
| I1112 14:45:45.514504 17193 solver.cpp:228] Iteration 197000, loss = 0.00573645 | |
| I1112 14:45:45.514578 17193 solver.cpp:244] Train net output #0: loss = 0.00573838 (* 1 = 0.00573838 loss) | |
| I1112 14:45:45.514593 17193 sgd_solver.cpp:106] Iteration 197000, lr = 3.57577e-06 | |
| I1112 14:47:55.910053 17193 solver.cpp:228] Iteration 198000, loss = 0.0153792 | |
| I1112 14:47:55.910140 17193 solver.cpp:244] Train net output #0: loss = 0.0153811 (* 1 = 0.0153811 loss) | |
| I1112 14:47:55.910156 17193 sgd_solver.cpp:106] Iteration 198000, lr = 3.56222e-06 | |
| I1112 14:50:06.333660 17193 solver.cpp:228] Iteration 199000, loss = 0.000937186 | |
| I1112 14:50:06.333837 17193 solver.cpp:244] Train net output #0: loss = 0.000939126 (* 1 = 0.000939126 loss) | |
| I1112 14:50:06.333853 17193 sgd_solver.cpp:106] Iteration 199000, lr = 3.54878e-06 | |
| I1112 14:52:16.605937 17193 solver.cpp:454] Snapshotting to binary proto file snapshots/snap__iter_200000.caffemodel | |
| I1112 14:52:44.548648 17193 sgd_solver.cpp:273] Snapshotting solver state to binary proto file snapshots/snap__iter_200000.solverstate | |
| I1112 14:52:45.106608 17193 solver.cpp:317] Iteration 200000, loss = 0.00837767 | |
| I1112 14:52:45.106654 17193 solver.cpp:337] Iteration 200000, Testing net (#0) | |
| I1112 14:52:45.106667 17193 net.cpp:693] Ignoring source layer drop1 | |
| I1112 14:52:45.106676 17193 net.cpp:693] Ignoring source layer prob | |
| I1112 14:52:52.327410 17193 solver.cpp:404] Test net output #0: accuracy = 0.99872 | |
| I1112 14:52:52.327476 17193 solver.cpp:322] Optimization Done. | |
| I1112 14:52:52.327486 17193 caffe.cpp:254] Optimization Done. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment