tail -500 /var/log/bozenka/retrain.log
2017-11-12 22:11:40.179934: Step 48530: Validation accuracy = 88.0% (N=100)
2017-11-12 22:11:40.939846: Step 48540: Train accuracy = 96.0%
2017-11-12 22:11:40.939912: Step 48540: Cross entropy = 0.148684
2017-11-12 22:11:41.008616: Step 48540: Validation accuracy = 75.0% (N=100)
2017-11-12 22:11:41.722998: Step 48550: Train accuracy = 97.0%
2017-11-12 22:11:41.723060: Step 48550: Cross entropy = 0.169537
2017-11-12 22:11:41.802840: Step 48550: Validation accuracy = 84.0% (N=100)
2017-11-12 22:11:42.530338: Step 48560: Train accuracy = 98.0%
2017-11-12 22:11:42.530399: Step 48560: Cross entropy = 0.127942
2017-11-12 22:11:42.597409: Step 48560: Validation accuracy = 81.0% (N=100)
2017-11-12 22:11:43.343959: Step 48570: Train accuracy = 95.0%
2017-11-12 22:11:43.344044: Step 48570: Cross entropy = 0.146929
2017-11-12 22:11:43.413488: Step 48570: Validation accuracy = 85.0% (N=100)
2017-11-12 22:11:44.165725: Step 48580: Train accuracy = 99.0%
2017-11-12 22:11:44.165800: Step 48580: Cross entropy = 0.133532
2017-11-12 22:11:44.240354: Step 48580: Validation accuracy = 84.0% (N=100)
2017-11-12 22:11:44.969711: Step 48590: Train accuracy = 94.0%
2017-11-12 22:11:44.969774: Step 48590: Cross entropy = 0.180877
2017-11-12 22:11:45.038857: Step 48590: Validation accuracy = 83.0% (N=100)
2017-11-12 22:11:45.725078: Step 48600: Train accuracy = 97.0%
2017-11-12 22:11:45.725157: Step 48600: Cross entropy = 0.140023
2017-11-12 22:11:45.793820: Step 48600: Validation accuracy = 83.0% (N=100)
2017-11-12 22:11:46.461888: Step 48610: Train accuracy = 97.0%
2017-11-12 22:11:46.461949: Step 48610: Cross entropy = 0.151413
2017-11-12 22:11:46.528055: Step 48610: Validation accuracy = 80.0% (N=100)
2017-11-12 22:11:47.158943: Step 48620: Train accuracy = 98.0%
2017-11-12 22:11:47.159004: Step 48620: Cross entropy = 0.151557
2017-11-12 22:11:47.221363: Step 48620: Validation accuracy = 88.0% (N=100)
2017-11-12 22:11:47.849994: Step 48630: Train accuracy = 95.0%
2017-11-12 22:11:47.850054: Step 48630: Cross entropy = 0.164243
2017-11-12 22:11:47.906702: Step 48630: Validation accuracy = 81.0% (N=100)
2017-11-12 22:11:48.559002: Step 48640: Train accuracy = 97.0%
2017-11-12 22:11:48.559064: Step 48640: Cross entropy = 0.132535
2017-11-12 22:11:48.619729: Step 48640: Validation accuracy = 83.0% (N=100)
2017-11-12 22:11:49.244098: Step 48650: Train accuracy = 96.0%
2017-11-12 22:11:49.244156: Step 48650: Cross entropy = 0.132258
2017-11-12 22:11:49.308144: Step 48650: Validation accuracy = 73.0% (N=100)
2017-11-12 22:11:49.919721: Step 48660: Train accuracy = 98.0%
2017-11-12 22:11:49.919778: Step 48660: Cross entropy = 0.113891
2017-11-12 22:11:49.980779: Step 48660: Validation accuracy = 88.0% (N=100)
2017-11-12 22:11:50.705009: Step 48670: Train accuracy = 100.0%
2017-11-12 22:11:50.705065: Step 48670: Cross entropy = 0.113341
2017-11-12 22:11:50.769047: Step 48670: Validation accuracy = 78.0% (N=100)
2017-11-12 22:11:51.393946: Step 48680: Train accuracy = 98.0%
2017-11-12 22:11:51.394009: Step 48680: Cross entropy = 0.113825
2017-11-12 22:11:51.461387: Step 48680: Validation accuracy = 91.0% (N=100)
2017-11-12 22:11:52.131109: Step 48690: Train accuracy = 99.0%
2017-11-12 22:11:52.131196: Step 48690: Cross entropy = 0.133232
2017-11-12 22:11:52.199544: Step 48690: Validation accuracy = 77.0% (N=100)
2017-11-12 22:11:52.840897: Step 48700: Train accuracy = 99.0%
2017-11-12 22:11:52.840960: Step 48700: Cross entropy = 0.107126
2017-11-12 22:11:52.904743: Step 48700: Validation accuracy = 83.0% (N=100)
2017-11-12 22:11:53.607517: Step 48710: Train accuracy = 98.0%
2017-11-12 22:11:53.607577: Step 48710: Cross entropy = 0.135007
2017-11-12 22:11:53.688124: Step 48710: Validation accuracy = 76.0% (N=100)
2017-11-12 22:11:54.368630: Step 48720: Train accuracy = 97.0%
2017-11-12 22:11:54.368687: Step 48720: Cross entropy = 0.127132
2017-11-12 22:11:54.427418: Step 48720: Validation accuracy = 84.0% (N=100)
2017-11-12 22:11:55.080581: Step 48730: Train accuracy = 99.0%
2017-11-12 22:11:55.080633: Step 48730: Cross entropy = 0.133716
2017-11-12 22:11:55.138318: Step 48730: Validation accuracy = 82.0% (N=100)
2017-11-12 22:11:55.856217: Step 48740: Train accuracy = 98.0%
2017-11-12 22:11:55.856330: Step 48740: Cross entropy = 0.143541
2017-11-12 22:11:55.932555: Step 48740: Validation accuracy = 81.0% (N=100)
2017-11-12 22:11:56.560355: Step 48750: Train accuracy = 97.0%
2017-11-12 22:11:56.560410: Step 48750: Cross entropy = 0.146033
2017-11-12 22:11:56.623445: Step 48750: Validation accuracy = 84.0% (N=100)
2017-11-12 22:11:57.318973: Step 48760: Train accuracy = 97.0%
2017-11-12 22:11:57.319038: Step 48760: Cross entropy = 0.133515
2017-11-12 22:11:57.381119: Step 48760: Validation accuracy = 85.0% (N=100)
2017-11-12 22:11:58.021407: Step 48770: Train accuracy = 99.0%
2017-11-12 22:11:58.021509: Step 48770: Cross entropy = 0.129193
2017-11-12 22:11:58.093141: Step 48770: Validation accuracy = 83.0% (N=100)
2017-11-12 22:11:58.811724: Step 48780: Train accuracy = 97.0%
2017-11-12 22:11:58.811820: Step 48780: Cross entropy = 0.109993
2017-11-12 22:11:58.885160: Step 48780: Validation accuracy = 84.0% (N=100)
2017-11-12 22:11:59.618142: Step 48790: Train accuracy = 95.0%
2017-11-12 22:11:59.618217: Step 48790: Cross entropy = 0.151739
2017-11-12 22:11:59.682598: Step 48790: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:00.384382: Step 48800: Train accuracy = 95.0%
2017-11-12 22:12:00.384432: Step 48800: Cross entropy = 0.169828
2017-11-12 22:12:00.452299: Step 48800: Validation accuracy = 76.0% (N=100)
2017-11-12 22:12:01.150559: Step 48810: Train accuracy = 97.0%
2017-11-12 22:12:01.150615: Step 48810: Cross entropy = 0.125921
2017-11-12 22:12:01.213192: Step 48810: Validation accuracy = 85.0% (N=100)
2017-11-12 22:12:01.915329: Step 48820: Train accuracy = 97.0%
2017-11-12 22:12:01.915384: Step 48820: Cross entropy = 0.181313
2017-11-12 22:12:01.981918: Step 48820: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:02.659337: Step 48830: Train accuracy = 99.0%
2017-11-12 22:12:02.659393: Step 48830: Cross entropy = 0.108951
2017-11-12 22:12:02.721062: Step 48830: Validation accuracy = 73.0% (N=100)
2017-11-12 22:12:03.411513: Step 48840: Train accuracy = 98.0%
2017-11-12 22:12:03.411626: Step 48840: Cross entropy = 0.119821
2017-11-12 22:12:03.482874: Step 48840: Validation accuracy = 75.0% (N=100)
2017-11-12 22:12:04.163176: Step 48850: Train accuracy = 97.0%
2017-11-12 22:12:04.163281: Step 48850: Cross entropy = 0.122728
2017-11-12 22:12:04.225419: Step 48850: Validation accuracy = 78.0% (N=100)
2017-11-12 22:12:04.860129: Step 48860: Train accuracy = 95.0%
2017-11-12 22:12:04.860190: Step 48860: Cross entropy = 0.189024
2017-11-12 22:12:04.923616: Step 48860: Validation accuracy = 74.0% (N=100)
2017-11-12 22:12:05.649742: Step 48870: Train accuracy = 98.0%
2017-11-12 22:12:05.649838: Step 48870: Cross entropy = 0.146287
2017-11-12 22:12:05.723555: Step 48870: Validation accuracy = 76.0% (N=100)
2017-11-12 22:12:06.381501: Step 48880: Train accuracy = 96.0%
2017-11-12 22:12:06.381560: Step 48880: Cross entropy = 0.168438
2017-11-12 22:12:06.459090: Step 48880: Validation accuracy = 73.0% (N=100)
2017-11-12 22:12:07.161353: Step 48890: Train accuracy = 98.0%
2017-11-12 22:12:07.161429: Step 48890: Cross entropy = 0.128761
2017-11-12 22:12:07.228383: Step 48890: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:07.894165: Step 48900: Train accuracy = 96.0%
2017-11-12 22:12:07.894240: Step 48900: Cross entropy = 0.139339
2017-11-12 22:12:07.967329: Step 48900: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:08.718799: Step 48910: Train accuracy = 98.0%
2017-11-12 22:12:08.718928: Step 48910: Cross entropy = 0.123057
2017-11-12 22:12:08.794410: Step 48910: Validation accuracy = 87.0% (N=100)
2017-11-12 22:12:09.446987: Step 48920: Train accuracy = 98.0%
2017-11-12 22:12:09.447072: Step 48920: Cross entropy = 0.109389
2017-11-12 22:12:09.517209: Step 48920: Validation accuracy = 79.0% (N=100)
2017-11-12 22:12:10.202898: Step 48930: Train accuracy = 97.0%
2017-11-12 22:12:10.202963: Step 48930: Cross entropy = 0.142757
2017-11-12 22:12:10.272063: Step 48930: Validation accuracy = 82.0% (N=100)
2017-11-12 22:12:11.004170: Step 48940: Train accuracy = 98.0%
2017-11-12 22:12:11.004256: Step 48940: Cross entropy = 0.153155
2017-11-12 22:12:11.068749: Step 48940: Validation accuracy = 78.0% (N=100)
2017-11-12 22:12:11.712042: Step 48950: Train accuracy = 98.0%
2017-11-12 22:12:11.712098: Step 48950: Cross entropy = 0.155999
2017-11-12 22:12:11.775068: Step 48950: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:12.464382: Step 48960: Train accuracy = 97.0%
2017-11-12 22:12:12.464437: Step 48960: Cross entropy = 0.147063
2017-11-12 22:12:12.529552: Step 48960: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:13.196372: Step 48970: Train accuracy = 98.0%
2017-11-12 22:12:13.196430: Step 48970: Cross entropy = 0.104813
2017-11-12 22:12:13.264725: Step 48970: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:13.950512: Step 48980: Train accuracy = 98.0%
2017-11-12 22:12:13.950571: Step 48980: Cross entropy = 0.123435
2017-11-12 22:12:14.020122: Step 48980: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:14.755097: Step 48990: Train accuracy = 96.0%
2017-11-12 22:12:14.755159: Step 48990: Cross entropy = 0.155526
2017-11-12 22:12:14.830784: Step 48990: Validation accuracy = 78.0% (N=100)
2017-11-12 22:12:15.549919: Step 49000: Train accuracy = 96.0%
2017-11-12 22:12:15.549981: Step 49000: Cross entropy = 0.135980
2017-11-12 22:12:15.626077: Step 49000: Validation accuracy = 80.0% (N=100)
2017-11-12 22:12:16.262044: Step 49010: Train accuracy = 98.0%
2017-11-12 22:12:16.262102: Step 49010: Cross entropy = 0.117121
2017-11-12 22:12:16.321722: Step 49010: Validation accuracy = 85.0% (N=100)
2017-11-12 22:12:17.024310: Step 49020: Train accuracy = 96.0%
2017-11-12 22:12:17.024368: Step 49020: Cross entropy = 0.134831
2017-11-12 22:12:17.090307: Step 49020: Validation accuracy = 88.0% (N=100)
2017-11-12 22:12:17.701482: Step 49030: Train accuracy = 98.0%
2017-11-12 22:12:17.701532: Step 49030: Cross entropy = 0.136366
2017-11-12 22:12:17.764622: Step 49030: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:18.434221: Step 49040: Train accuracy = 97.0%
2017-11-12 22:12:18.434283: Step 49040: Cross entropy = 0.162917
2017-11-12 22:12:18.504694: Step 49040: Validation accuracy = 85.0% (N=100)
2017-11-12 22:12:19.133947: Step 49050: Train accuracy = 98.0%
2017-11-12 22:12:19.134007: Step 49050: Cross entropy = 0.127813
2017-11-12 22:12:19.195789: Step 49050: Validation accuracy = 80.0% (N=100)
2017-11-12 22:12:19.826896: Step 49060: Train accuracy = 98.0%
2017-11-12 22:12:19.826957: Step 49060: Cross entropy = 0.135781
2017-11-12 22:12:19.887336: Step 49060: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:20.613739: Step 49070: Train accuracy = 97.0%
2017-11-12 22:12:20.613796: Step 49070: Cross entropy = 0.127294
2017-11-12 22:12:20.698923: Step 49070: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:21.379756: Step 49080: Train accuracy = 98.0%
2017-11-12 22:12:21.379814: Step 49080: Cross entropy = 0.101884
2017-11-12 22:12:21.449198: Step 49080: Validation accuracy = 89.0% (N=100)
2017-11-12 22:12:22.121334: Step 49090: Train accuracy = 97.0%
2017-11-12 22:12:22.121393: Step 49090: Cross entropy = 0.123114
2017-11-12 22:12:22.202412: Step 49090: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:22.852270: Step 49100: Train accuracy = 98.0%
2017-11-12 22:12:22.852342: Step 49100: Cross entropy = 0.134860
2017-11-12 22:12:22.914864: Step 49100: Validation accuracy = 82.0% (N=100)
2017-11-12 22:12:23.617576: Step 49110: Train accuracy = 100.0%
2017-11-12 22:12:23.617638: Step 49110: Cross entropy = 0.087226
2017-11-12 22:12:23.694103: Step 49110: Validation accuracy = 82.0% (N=100)
2017-11-12 22:12:24.380589: Step 49120: Train accuracy = 99.0%
2017-11-12 22:12:24.380644: Step 49120: Cross entropy = 0.110012
2017-11-12 22:12:24.442605: Step 49120: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:25.115573: Step 49130: Train accuracy = 98.0%
2017-11-12 22:12:25.115630: Step 49130: Cross entropy = 0.122157
2017-11-12 22:12:25.191238: Step 49130: Validation accuracy = 79.0% (N=100)
2017-11-12 22:12:25.901261: Step 49140: Train accuracy = 96.0%
2017-11-12 22:12:25.901313: Step 49140: Cross entropy = 0.124339
2017-11-12 22:12:25.961074: Step 49140: Validation accuracy = 80.0% (N=100)
2017-11-12 22:12:26.590433: Step 49150: Train accuracy = 96.0%
2017-11-12 22:12:26.590518: Step 49150: Cross entropy = 0.130777
2017-11-12 22:12:26.654458: Step 49150: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:27.331677: Step 49160: Train accuracy = 98.0%
2017-11-12 22:12:27.331733: Step 49160: Cross entropy = 0.129351
2017-11-12 22:12:27.390957: Step 49160: Validation accuracy = 88.0% (N=100)
2017-11-12 22:12:28.041128: Step 49170: Train accuracy = 95.0%
2017-11-12 22:12:28.041191: Step 49170: Cross entropy = 0.136441
2017-11-12 22:12:28.107913: Step 49170: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:28.778655: Step 49180: Train accuracy = 99.0%
2017-11-12 22:12:28.778711: Step 49180: Cross entropy = 0.130359
2017-11-12 22:12:28.845725: Step 49180: Validation accuracy = 90.0% (N=100)
2017-11-12 22:12:29.600972: Step 49190: Train accuracy = 99.0%
2017-11-12 22:12:29.601030: Step 49190: Cross entropy = 0.130338
2017-11-12 22:12:29.661916: Step 49190: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:30.363385: Step 49200: Train accuracy = 98.0%
2017-11-12 22:12:30.363438: Step 49200: Cross entropy = 0.168286
2017-11-12 22:12:30.437199: Step 49200: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:31.144463: Step 49210: Train accuracy = 100.0%
2017-11-12 22:12:31.144517: Step 49210: Cross entropy = 0.113167
2017-11-12 22:12:31.214524: Step 49210: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:31.878461: Step 49220: Train accuracy = 97.0%
2017-11-12 22:12:31.878515: Step 49220: Cross entropy = 0.158251
2017-11-12 22:12:31.937043: Step 49220: Validation accuracy = 77.0% (N=100)
2017-11-12 22:12:32.564736: Step 49230: Train accuracy = 97.0%
2017-11-12 22:12:32.564806: Step 49230: Cross entropy = 0.159037
2017-11-12 22:12:32.627571: Step 49230: Validation accuracy = 85.0% (N=100)
2017-11-12 22:12:33.294461: Step 49240: Train accuracy = 95.0%
2017-11-12 22:12:33.294519: Step 49240: Cross entropy = 0.152758
2017-11-12 22:12:33.369131: Step 49240: Validation accuracy = 82.0% (N=100)
2017-11-12 22:12:34.077397: Step 49250: Train accuracy = 100.0%
2017-11-12 22:12:34.077455: Step 49250: Cross entropy = 0.133827
2017-11-12 22:12:34.141691: Step 49250: Validation accuracy = 75.0% (N=100)
2017-11-12 22:12:34.786979: Step 49260: Train accuracy = 96.0%
2017-11-12 22:12:34.787040: Step 49260: Cross entropy = 0.152693
2017-11-12 22:12:34.847279: Step 49260: Validation accuracy = 85.0% (N=100)
2017-11-12 22:12:35.540791: Step 49270: Train accuracy = 96.0%
2017-11-12 22:12:35.540850: Step 49270: Cross entropy = 0.135034
2017-11-12 22:12:35.618172: Step 49270: Validation accuracy = 82.0% (N=100)
2017-11-12 22:12:36.289424: Step 49280: Train accuracy = 97.0%
2017-11-12 22:12:36.289479: Step 49280: Cross entropy = 0.130112
2017-11-12 22:12:36.348653: Step 49280: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:37.026779: Step 49290: Train accuracy = 96.0%
2017-11-12 22:12:37.026873: Step 49290: Cross entropy = 0.135216
2017-11-12 22:12:37.098400: Step 49290: Validation accuracy = 80.0% (N=100)
2017-11-12 22:12:37.740159: Step 49300: Train accuracy = 98.0%
2017-11-12 22:12:37.740262: Step 49300: Cross entropy = 0.121298
2017-11-12 22:12:37.796907: Step 49300: Validation accuracy = 77.0% (N=100)
2017-11-12 22:12:38.503393: Step 49310: Train accuracy = 97.0%
2017-11-12 22:12:38.503471: Step 49310: Cross entropy = 0.138424
2017-11-12 22:12:38.581076: Step 49310: Validation accuracy = 88.0% (N=100)
2017-11-12 22:12:39.354544: Step 49320: Train accuracy = 97.0%
2017-11-12 22:12:39.354600: Step 49320: Cross entropy = 0.143137
2017-11-12 22:12:39.430308: Step 49320: Validation accuracy = 78.0% (N=100)
2017-11-12 22:12:40.177033: Step 49330: Train accuracy = 95.0%
2017-11-12 22:12:40.177097: Step 49330: Cross entropy = 0.175720
2017-11-12 22:12:40.251177: Step 49330: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:41.045658: Step 49340: Train accuracy = 99.0%
2017-11-12 22:12:41.045714: Step 49340: Cross entropy = 0.108893
2017-11-12 22:12:41.122831: Step 49340: Validation accuracy = 79.0% (N=100)
2017-11-12 22:12:41.894042: Step 49350: Train accuracy = 98.0%
2017-11-12 22:12:41.894110: Step 49350: Cross entropy = 0.114504
2017-11-12 22:12:41.971036: Step 49350: Validation accuracy = 88.0% (N=100)
2017-11-12 22:12:42.712926: Step 49360: Train accuracy = 96.0%
2017-11-12 22:12:42.712990: Step 49360: Cross entropy = 0.150874
2017-11-12 22:12:42.780149: Step 49360: Validation accuracy = 76.0% (N=100)
2017-11-12 22:12:43.534468: Step 49370: Train accuracy = 97.0%
2017-11-12 22:12:43.534534: Step 49370: Cross entropy = 0.141518
2017-11-12 22:12:43.607020: Step 49370: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:44.383535: Step 49380: Train accuracy = 98.0%
2017-11-12 22:12:44.383611: Step 49380: Cross entropy = 0.141415
2017-11-12 22:12:44.463020: Step 49380: Validation accuracy = 80.0% (N=100)
2017-11-12 22:12:45.187664: Step 49390: Train accuracy = 99.0%
2017-11-12 22:12:45.187720: Step 49390: Cross entropy = 0.135613
2017-11-12 22:12:45.257429: Step 49390: Validation accuracy = 74.0% (N=100)
2017-11-12 22:12:45.939000: Step 49400: Train accuracy = 97.0%
2017-11-12 22:12:45.939083: Step 49400: Cross entropy = 0.115414
2017-11-12 22:12:45.999458: Step 49400: Validation accuracy = 86.0% (N=100)
2017-11-12 22:12:46.679436: Step 49410: Train accuracy = 98.0%
2017-11-12 22:12:46.679487: Step 49410: Cross entropy = 0.116749
2017-11-12 22:12:46.744362: Step 49410: Validation accuracy = 90.0% (N=100)
2017-11-12 22:12:47.407239: Step 49420: Train accuracy = 98.0%
2017-11-12 22:12:47.407292: Step 49420: Cross entropy = 0.120579
2017-11-12 22:12:47.467416: Step 49420: Validation accuracy = 77.0% (N=100)
2017-11-12 22:12:48.097760: Step 49430: Train accuracy = 96.0%
2017-11-12 22:12:48.097808: Step 49430: Cross entropy = 0.151138
2017-11-12 22:12:48.163008: Step 49430: Validation accuracy = 74.0% (N=100)
2017-11-12 22:12:48.865488: Step 49440: Train accuracy = 98.0%
2017-11-12 22:12:48.865544: Step 49440: Cross entropy = 0.097154
2017-11-12 22:12:48.925642: Step 49440: Validation accuracy = 83.0% (N=100)
2017-11-12 22:12:49.570865: Step 49450: Train accuracy = 100.0%
2017-11-12 22:12:49.570964: Step 49450: Cross entropy = 0.113068
2017-11-12 22:12:49.630176: Step 49450: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:50.283571: Step 49460: Train accuracy = 100.0%
2017-11-12 22:12:50.283626: Step 49460: Cross entropy = 0.106607
2017-11-12 22:12:50.352209: Step 49460: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:51.051746: Step 49470: Train accuracy = 97.0%
2017-11-12 22:12:51.051801: Step 49470: Cross entropy = 0.136575
2017-11-12 22:12:51.113429: Step 49470: Validation accuracy = 79.0% (N=100)
2017-11-12 22:12:51.812341: Step 49480: Train accuracy = 96.0%
2017-11-12 22:12:51.812397: Step 49480: Cross entropy = 0.140645
2017-11-12 22:12:51.871360: Step 49480: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:52.536680: Step 49490: Train accuracy = 98.0%
2017-11-12 22:12:52.536729: Step 49490: Cross entropy = 0.117187
2017-11-12 22:12:52.602960: Step 49490: Validation accuracy = 79.0% (N=100)
2017-11-12 22:12:53.279133: Step 49500: Train accuracy = 97.0%
2017-11-12 22:12:53.279211: Step 49500: Cross entropy = 0.126923
2017-11-12 22:12:53.340061: Step 49500: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:54.079644: Step 49510: Train accuracy = 95.0%
2017-11-12 22:12:54.079699: Step 49510: Cross entropy = 0.159582
2017-11-12 22:12:54.143026: Step 49510: Validation accuracy = 76.0% (N=100)
2017-11-12 22:12:54.784824: Step 49520: Train accuracy = 96.0%
2017-11-12 22:12:54.784891: Step 49520: Cross entropy = 0.151112
2017-11-12 22:12:54.853138: Step 49520: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:55.553925: Step 49530: Train accuracy = 98.0%
2017-11-12 22:12:55.553988: Step 49530: Cross entropy = 0.127737
2017-11-12 22:12:55.628362: Step 49530: Validation accuracy = 87.0% (N=100)
2017-11-12 22:12:56.315022: Step 49540: Train accuracy = 100.0%
2017-11-12 22:12:56.315078: Step 49540: Cross entropy = 0.099836
2017-11-12 22:12:56.384729: Step 49540: Validation accuracy = 88.0% (N=100)
2017-11-12 22:12:57.039637: Step 49550: Train accuracy = 99.0%
2017-11-12 22:12:57.039695: Step 49550: Cross entropy = 0.148036
2017-11-12 22:12:57.106318: Step 49550: Validation accuracy = 84.0% (N=100)
2017-11-12 22:12:57.744140: Step 49560: Train accuracy = 95.0%
2017-11-12 22:12:57.744200: Step 49560: Cross entropy = 0.158472
2017-11-12 22:12:57.801722: Step 49560: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:58.481786: Step 49570: Train accuracy = 96.0%
2017-11-12 22:12:58.481844: Step 49570: Cross entropy = 0.140798
2017-11-12 22:12:58.547972: Step 49570: Validation accuracy = 81.0% (N=100)
2017-11-12 22:12:59.302593: Step 49580: Train accuracy = 100.0%
2017-11-12 22:12:59.302648: Step 49580: Cross entropy = 0.139121
2017-11-12 22:12:59.368142: Step 49580: Validation accuracy = 82.0% (N=100)
2017-11-12 22:13:00.096161: Step 49590: Train accuracy = 96.0%
2017-11-12 22:13:00.096241: Step 49590: Cross entropy = 0.162918
2017-11-12 22:13:00.165262: Step 49590: Validation accuracy = 79.0% (N=100)
2017-11-12 22:13:00.863655: Step 49600: Train accuracy = 97.0%
2017-11-12 22:13:00.863715: Step 49600: Cross entropy = 0.170295
2017-11-12 22:13:00.933608: Step 49600: Validation accuracy = 90.0% (N=100)
2017-11-12 22:13:01.588508: Step 49610: Train accuracy = 97.0%
2017-11-12 22:13:01.588569: Step 49610: Cross entropy = 0.177203
2017-11-12 22:13:01.657203: Step 49610: Validation accuracy = 72.0% (N=100)
2017-11-12 22:13:02.336833: Step 49620: Train accuracy = 98.0%
2017-11-12 22:13:02.336940: Step 49620: Cross entropy = 0.119102
2017-11-12 22:13:02.398664: Step 49620: Validation accuracy = 81.0% (N=100)
2017-11-12 22:13:03.043789: Step 49630: Train accuracy = 96.0%
2017-11-12 22:13:03.043856: Step 49630: Cross entropy = 0.173210
2017-11-12 22:13:03.110693: Step 49630: Validation accuracy = 85.0% (N=100)
2017-11-12 22:13:03.808511: Step 49640: Train accuracy = 93.0%
2017-11-12 22:13:03.808569: Step 49640: Cross entropy = 0.170365
2017-11-12 22:13:03.873236: Step 49640: Validation accuracy = 77.0% (N=100)
2017-11-12 22:13:04.533333: Step 49650: Train accuracy = 98.0%
2017-11-12 22:13:04.533391: Step 49650: Cross entropy = 0.131990
2017-11-12 22:13:04.599453: Step 49650: Validation accuracy = 90.0% (N=100)
2017-11-12 22:13:05.261017: Step 49660: Train accuracy = 97.0%
2017-11-12 22:13:05.261081: Step 49660: Cross entropy = 0.143702
2017-11-12 22:13:05.330464: Step 49660: Validation accuracy = 77.0% (N=100)
2017-11-12 22:13:06.067927: Step 49670: Train accuracy = 99.0%
2017-11-12 22:13:06.068033: Step 49670: Cross entropy = 0.142035
2017-11-12 22:13:06.130737: Step 49670: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:06.790872: Step 49680: Train accuracy = 96.0%
2017-11-12 22:13:06.790949: Step 49680: Cross entropy = 0.147010
2017-11-12 22:13:06.853189: Step 49680: Validation accuracy = 84.0% (N=100)
2017-11-12 22:13:07.536463: Step 49690: Train accuracy = 99.0%
2017-11-12 22:13:07.536518: Step 49690: Cross entropy = 0.135149
2017-11-12 22:13:07.607194: Step 49690: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:08.311954: Step 49700: Train accuracy = 97.0%
2017-11-12 22:13:08.312054: Step 49700: Cross entropy = 0.120421
2017-11-12 22:13:08.378751: Step 49700: Validation accuracy = 77.0% (N=100)
2017-11-12 22:13:09.140079: Step 49710: Train accuracy = 98.0%
2017-11-12 22:13:09.140141: Step 49710: Cross entropy = 0.153702
2017-11-12 22:13:09.201946: Step 49710: Validation accuracy = 84.0% (N=100)
2017-11-12 22:13:09.875651: Step 49720: Train accuracy = 98.0%
2017-11-12 22:13:09.875729: Step 49720: Cross entropy = 0.136479
2017-11-12 22:13:09.955557: Step 49720: Validation accuracy = 78.0% (N=100)
2017-11-12 22:13:10.689109: Step 49730: Train accuracy = 98.0%
2017-11-12 22:13:10.689172: Step 49730: Cross entropy = 0.147626
2017-11-12 22:13:10.754358: Step 49730: Validation accuracy = 85.0% (N=100)
2017-11-12 22:13:11.429316: Step 49740: Train accuracy = 92.0%
2017-11-12 22:13:11.429366: Step 49740: Cross entropy = 0.158749
2017-11-12 22:13:11.491021: Step 49740: Validation accuracy = 88.0% (N=100)
2017-11-12 22:13:12.178012: Step 49750: Train accuracy = 98.0%
2017-11-12 22:13:12.178082: Step 49750: Cross entropy = 0.135615
2017-11-12 22:13:12.239127: Step 49750: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:12.871584: Step 49760: Train accuracy = 95.0%
2017-11-12 22:13:12.871635: Step 49760: Cross entropy = 0.179336
2017-11-12 22:13:12.936048: Step 49760: Validation accuracy = 85.0% (N=100)
2017-11-12 22:13:13.632238: Step 49770: Train accuracy = 96.0%
2017-11-12 22:13:13.632341: Step 49770: Cross entropy = 0.152591
2017-11-12 22:13:13.696954: Step 49770: Validation accuracy = 86.0% (N=100)
2017-11-12 22:13:14.437519: Step 49780: Train accuracy = 100.0%
2017-11-12 22:13:14.437574: Step 49780: Cross entropy = 0.112428
2017-11-12 22:13:14.506767: Step 49780: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:15.234102: Step 49790: Train accuracy = 97.0%
2017-11-12 22:13:15.234165: Step 49790: Cross entropy = 0.157685
2017-11-12 22:13:15.294741: Step 49790: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:16.002123: Step 49800: Train accuracy = 95.0%
2017-11-12 22:13:16.002178: Step 49800: Cross entropy = 0.124382
2017-11-12 22:13:16.067692: Step 49800: Validation accuracy = 91.0% (N=100)
2017-11-12 22:13:16.771845: Step 49810: Train accuracy = 98.0%
2017-11-12 22:13:16.772062: Step 49810: Cross entropy = 0.133184
2017-11-12 22:13:16.835610: Step 49810: Validation accuracy = 78.0% (N=100)
2017-11-12 22:13:17.472590: Step 49820: Train accuracy = 99.0%
2017-11-12 22:13:17.472645: Step 49820: Cross entropy = 0.092537
2017-11-12 22:13:17.533778: Step 49820: Validation accuracy = 83.0% (N=100)
2017-11-12 22:13:18.158489: Step 49830: Train accuracy = 98.0%
2017-11-12 22:13:18.158540: Step 49830: Cross entropy = 0.104184
2017-11-12 22:13:18.234034: Step 49830: Validation accuracy = 88.0% (N=100)
2017-11-12 22:13:18.895756: Step 49840: Train accuracy = 94.0%
2017-11-12 22:13:18.895850: Step 49840: Cross entropy = 0.165328
2017-11-12 22:13:18.959667: Step 49840: Validation accuracy = 89.0% (N=100)
2017-11-12 22:13:19.594614: Step 49850: Train accuracy = 97.0%
2017-11-12 22:13:19.594667: Step 49850: Cross entropy = 0.139358
2017-11-12 22:13:19.656434: Step 49850: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:20.344592: Step 49860: Train accuracy = 94.0%
2017-11-12 22:13:20.344656: Step 49860: Cross entropy = 0.186072
2017-11-12 22:13:20.420480: Step 49860: Validation accuracy = 82.0% (N=100)
2017-11-12 22:13:21.133791: Step 49870: Train accuracy = 96.0%
2017-11-12 22:13:21.133844: Step 49870: Cross entropy = 0.142697
2017-11-12 22:13:21.196672: Step 49870: Validation accuracy = 77.0% (N=100)
2017-11-12 22:13:21.873150: Step 49880: Train accuracy = 98.0%
2017-11-12 22:13:21.873210: Step 49880: Cross entropy = 0.113444
2017-11-12 22:13:21.934696: Step 49880: Validation accuracy = 77.0% (N=100)
2017-11-12 22:13:22.617427: Step 49890: Train accuracy = 98.0%
2017-11-12 22:13:22.617483: Step 49890: Cross entropy = 0.134863
2017-11-12 22:13:22.681578: Step 49890: Validation accuracy = 91.0% (N=100)
2017-11-12 22:13:23.409932: Step 49900: Train accuracy = 95.0%
2017-11-12 22:13:23.410030: Step 49900: Cross entropy = 0.173574
2017-11-12 22:13:23.479539: Step 49900: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:24.221389: Step 49910: Train accuracy = 97.0%
2017-11-12 22:13:24.221442: Step 49910: Cross entropy = 0.133961
2017-11-12 22:13:24.283323: Step 49910: Validation accuracy = 88.0% (N=100)
2017-11-12 22:13:24.962682: Step 49920: Train accuracy = 97.0%
2017-11-12 22:13:24.962761: Step 49920: Cross entropy = 0.133783
2017-11-12 22:13:25.034068: Step 49920: Validation accuracy = 82.0% (N=100)
2017-11-12 22:13:25.781345: Step 49930: Train accuracy = 97.0%
2017-11-12 22:13:25.781410: Step 49930: Cross entropy = 0.122591
2017-11-12 22:13:25.843088: Step 49930: Validation accuracy = 73.0% (N=100)
2017-11-12 22:13:26.474423: Step 49940: Train accuracy = 96.0%
2017-11-12 22:13:26.474478: Step 49940: Cross entropy = 0.125110
2017-11-12 22:13:26.535625: Step 49940: Validation accuracy = 80.0% (N=100)
2017-11-12 22:13:27.212067: Step 49950: Train accuracy = 93.0%
2017-11-12 22:13:27.212131: Step 49950: Cross entropy = 0.167996
2017-11-12 22:13:27.276505: Step 49950: Validation accuracy = 81.0% (N=100)
2017-11-12 22:13:27.918683: Step 49960: Train accuracy = 97.0%
2017-11-12 22:13:27.918740: Step 49960: Cross entropy = 0.131596
2017-11-12 22:13:27.986231: Step 49960: Validation accuracy = 85.0% (N=100)
2017-11-12 22:13:28.662887: Step 49970: Train accuracy = 97.0%
2017-11-12 22:13:28.663016: Step 49970: Cross entropy = 0.110505
2017-11-12 22:13:28.733108: Step 49970: Validation accuracy = 86.0% (N=100)
2017-11-12 22:13:29.469452: Step 49980: Train accuracy = 95.0%
2017-11-12 22:13:29.469523: Step 49980: Cross entropy = 0.152911
2017-11-12 22:13:29.545571: Step 49980: Validation accuracy = 93.0% (N=100)
2017-11-12 22:13:30.272868: Step 49990: Train accuracy = 98.0%
2017-11-12 22:13:30.273018: Step 49990: Cross entropy = 0.120465
2017-11-12 22:13:30.335808: Step 49990: Validation accuracy = 85.0% (N=100)
2017-11-12 22:13:30.951836: Step 49999: Train accuracy = 97.0%
2017-11-12 22:13:30.951909: Step 49999: Cross entropy = 0.147950
2017-11-12 22:13:31.019962: Step 49999: Validation accuracy = 78.0% (N=100)
Final test accuracy = 82.1% (N=308)
=== MISCLASSIFIED TEST IMAGES ===
                          library/zatazeno/strojetice_170610152137.jpg  polojasno
                             library/zatazeno/pribram_170612120003.jpg  polojasno
                          library/zatazeno/lysa_hora3_170612120009.jpg  polojasno
                            library/zatazeno/frydlant_170611130005.jpg  polojasno
                                 library/zatazeno/sec_170612040009.jpg  polojasno
                              library/zatazeno/vitkov_170612110009.jpg  polojasno
                             library/zatazeno/klatovy_170611040001.jpg  jasno
                             library/zatazeno/rymarov_170612100009.jpg  polojasno
                                library/zatazeno/brdy_170612100004.jpg  polojasno
                               library/zatazeno/polom_170612090008.jpg  polojasno
                           library/zatazeno/svratouch_170612080008.jpg  polojasno
                            library/zatazeno/frydlant_170611150007.jpg  polojasno
                      library/zatazeno/hradec_kralove_170611150009.jpg  polojasno
                               library/zatazeno/serak_170611180008.jpg  polojasno
                                library/jasno/belotin_170611090008.jpg  polojasno
                                  library/jasno/mrzky_170611030005.jpg  zatazeno
                                library/jasno/maruska_170611140011.jpg  polojasno
                              library/jasno/kralovice_170611160002.jpg  polojasno
                                library/jasno/klatovy_170611050003.jpg  zatazeno
                                 library/jasno/plzen2_170611020001.jpg  zatazeno
                                 library/jasno/bucina_170611050004.jpg  zatazeno
                                    library/jasno/sec_170611030007.jpg  zatazeno
                            library/jasno/labskabouda_170612110006.jpg  polojasno
                                   library/jasno/cheb_170611100000.jpg  polojasno
                               library/jasno/svatobor_170610160002.jpg  polojasno
                                   library/jasno/luka_170610160009.jpg  polojasno
                                library/jasno/paprsek_170610180011.jpg  polojasno
                                 library/jasno/bucina_170612110003.jpg  polojasno
                         library/polojasno/libinprach_170612070005.jpg  jasno
                          library/polojasno/kralovice_170611110002.jpg  jasno
                            library/polojasno/olomouc_170610170010.jpg  jasno
                            library/polojasno/medenec_170611030001.jpg  jasno
                            library/polojasno/medenec_170611020001.jpg  jasno
                       library/polojasno/ustinadlabem_170611110004.jpg  jasno
                         library/polojasno/kocelovice_170610143023.jpg  zatazeno
                   library/polojasno/ceske_budejovice_170611150006.jpg  zatazeno
                            library/polojasno/rymarov_170610152144.jpg  zatazeno
                             library/polojasno/vsetin_170612090010.jpg  jasno
                        library/polojasno/kucharovice_170611120008.jpg  jasno
                         library/polojasno/lucnibouda_170610152141.jpg  zatazeno
                           library/polojasno/frymburk_170611050006.jpg  jasno
                     library/polojasno/ostrava_poruba_170611170010.jpg  jasno
                               library/polojasno/lkmt_170612020008.jpg  jasno
                             library/polojasno/skalky_170610180010.jpg  jasno
                           library/polojasno/osoblaha_170612070010.jpg  zatazeno
                            library/polojasno/hejnice_170611150007.jpg  zatazeno
                         library/polojasno/kocelovice_170612090004.jpg  zatazeno
                           library/polojasno/vlkonice_170610160002.jpg  zatazeno
                           library/polojasno/klinovec_170612100002.jpg  zatazeno
                         library/polojasno/dyjakovice_170611130007.jpg  zatazeno
                            library/polojasno/paprsek_170611070010.jpg  jasno
                            library/polojasno/olomouc_170610160009.jpg  jasno
                              library/polojasno/dylen_170612090001.jpg  jasno
                     library/polojasno/ostrava_poruba_170612070010.jpg  jasno
                           library/polojasno/slamenka_170611070010.jpg  jasno
Converted 2 variables to const ops.