20170909深度学习solar测试日志

py@py-System-Product-Name:~/deep_learning/TFLearn/re$ python retrain.py –image_dir ~/20170513DP
Looking for images in ‘BadDP’
Looking for images in ‘GoodDP’
2017-09-09 16:26:23.319098: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:26:23.319118: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:26:23.319123: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:26:23.319127: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:26:23.319130: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:26:23.460825: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2017-09-09 16:26:23.461147: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties:
name: GeForce GTX 1080
major: 6 minor: 1 memoryClockRate (GHz) 1.8225
pciBusID 0000:01:00.0
Total memory: 7.92GiB
Free memory: 7.50GiB
2017-09-09 16:26:23.556993: W tensorflow/stream_executor/cuda/cuda_driver.cc:523] A non-primary context 0x3d03280 exists before initializing the StreamExecutor. We haven’t verified StreamExecutor works with that.
2017-09-09 16:26:23.557253: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2017-09-09 16:26:23.557556: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 1 with properties:
name: GeForce GTX 1080
major: 6 minor: 1 memoryClockRate (GHz) 1.8225
pciBusID 0000:02:00.0
Total memory: 7.92GiB
Free memory: 7.81GiB
2017-09-09 16:26:23.558064: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 1
2017-09-09 16:26:23.558072: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y Y
2017-09-09 16:26:23.558075: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 1: Y Y
2017-09-09 16:26:23.558081: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1080, pci bus id: 0000:01:00.0)
2017-09-09 16:26:23.558084: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 1080, pci bus id: 0000:02:00.0)
100 bottleneck files created.
200 bottleneck files created.
300 bottleneck files created.
400 bottleneck files created.
500 bottleneck files created.
600 bottleneck files created.
700 bottleneck files created.
800 bottleneck files created.
900 bottleneck files created.
2017-09-09 16:26:24.601367: Step 0: Train accuracy = 76.7%
2017-09-09 16:26:24.601412: Step 0: Cross entropy = 0.687446
2017-09-09 16:26:24.617348: Step 0: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:24.736530: Step 10: Train accuracy = 63.3%
2017-09-09 16:26:24.736571: Step 10: Cross entropy = 0.644534
2017-09-09 16:26:24.747895: Step 10: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:24.866573: Step 20: Train accuracy = 63.3%
2017-09-09 16:26:24.866614: Step 20: Cross entropy = 0.652286
2017-09-09 16:26:24.878019: Step 20: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:24.996162: Step 30: Train accuracy = 63.3%
2017-09-09 16:26:24.996204: Step 30: Cross entropy = 0.583036
2017-09-09 16:26:25.007631: Step 30: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:25.126408: Step 40: Train accuracy = 66.7%
2017-09-09 16:26:25.126449: Step 40: Cross entropy = 0.571332
2017-09-09 16:26:25.137916: Step 40: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:25.256038: Step 50: Train accuracy = 63.3%
2017-09-09 16:26:25.256079: Step 50: Cross entropy = 0.588173
2017-09-09 16:26:25.267469: Step 50: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:25.386110: Step 60: Train accuracy = 83.3%
2017-09-09 16:26:25.386152: Step 60: Cross entropy = 0.586895
2017-09-09 16:26:25.398382: Step 60: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:25.518094: Step 70: Train accuracy = 66.7%
2017-09-09 16:26:25.518137: Step 70: Cross entropy = 0.632733
2017-09-09 16:26:25.529698: Step 70: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:25.648355: Step 80: Train accuracy = 63.3%
2017-09-09 16:26:25.648396: Step 80: Cross entropy = 0.590801
2017-09-09 16:26:25.660017: Step 80: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:25.777880: Step 90: Train accuracy = 80.0%
2017-09-09 16:26:25.777921: Step 90: Cross entropy = 0.555562
2017-09-09 16:26:25.789417: Step 90: Validation accuracy = 50.0% (N=30)
2017-09-09 16:26:25.907431: Step 100: Train accuracy = 60.0%
2017-09-09 16:26:25.907474: Step 100: Cross entropy = 0.572725
2017-09-09 16:26:25.918983: Step 100: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:26.037122: Step 110: Train accuracy = 63.3%
2017-09-09 16:26:26.037163: Step 110: Cross entropy = 0.593043
2017-09-09 16:26:26.049443: Step 110: Validation accuracy = 46.7% (N=30)
2017-09-09 16:26:26.167256: Step 120: Train accuracy = 76.7%
2017-09-09 16:26:26.167299: Step 120: Cross entropy = 0.502749
2017-09-09 16:26:26.178719: Step 120: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:26.297025: Step 130: Train accuracy = 73.3%
2017-09-09 16:26:26.297067: Step 130: Cross entropy = 0.578728
2017-09-09 16:26:26.308500: Step 130: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:26.426064: Step 140: Train accuracy = 66.7%
2017-09-09 16:26:26.426106: Step 140: Cross entropy = 0.594375
2017-09-09 16:26:26.437862: Step 140: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:26.557825: Step 150: Train accuracy = 83.3%
2017-09-09 16:26:26.557867: Step 150: Cross entropy = 0.537235
2017-09-09 16:26:26.569293: Step 150: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:26.688136: Step 160: Train accuracy = 70.0%
2017-09-09 16:26:26.688179: Step 160: Cross entropy = 0.574994
2017-09-09 16:26:26.699615: Step 160: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:26.817651: Step 170: Train accuracy = 70.0%
2017-09-09 16:26:26.817692: Step 170: Cross entropy = 0.568421
2017-09-09 16:26:26.829255: Step 170: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:26.948506: Step 180: Train accuracy = 70.0%
2017-09-09 16:26:26.948548: Step 180: Cross entropy = 0.613707
2017-09-09 16:26:26.959964: Step 180: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:27.077317: Step 190: Train accuracy = 70.0%
2017-09-09 16:26:27.077358: Step 190: Cross entropy = 0.542539
2017-09-09 16:26:27.088867: Step 190: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:27.206485: Step 200: Train accuracy = 73.3%
2017-09-09 16:26:27.206527: Step 200: Cross entropy = 0.518869
2017-09-09 16:26:27.218167: Step 200: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:27.336186: Step 210: Train accuracy = 66.7%
2017-09-09 16:26:27.336227: Step 210: Cross entropy = 0.593715
2017-09-09 16:26:27.347652: Step 210: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:27.466238: Step 220: Train accuracy = 86.7%
2017-09-09 16:26:27.466279: Step 220: Cross entropy = 0.490490
2017-09-09 16:26:27.477720: Step 220: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:27.596539: Step 230: Train accuracy = 76.7%
2017-09-09 16:26:27.596581: Step 230: Cross entropy = 0.494022
2017-09-09 16:26:27.608150: Step 230: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:27.727769: Step 240: Train accuracy = 73.3%
2017-09-09 16:26:27.727816: Step 240: Cross entropy = 0.507795
2017-09-09 16:26:27.739168: Step 240: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:27.857602: Step 250: Train accuracy = 73.3%
2017-09-09 16:26:27.857644: Step 250: Cross entropy = 0.547760
2017-09-09 16:26:27.869138: Step 250: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:27.987221: Step 260: Train accuracy = 56.7%
2017-09-09 16:26:27.987267: Step 260: Cross entropy = 0.635385
2017-09-09 16:26:27.999418: Step 260: Validation accuracy = 46.7% (N=30)
2017-09-09 16:26:28.119015: Step 270: Train accuracy = 80.0%
2017-09-09 16:26:28.119057: Step 270: Cross entropy = 0.482785
2017-09-09 16:26:28.130487: Step 270: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:28.249289: Step 280: Train accuracy = 90.0%
2017-09-09 16:26:28.249330: Step 280: Cross entropy = 0.419106
2017-09-09 16:26:28.260937: Step 280: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:28.378847: Step 290: Train accuracy = 63.3%
2017-09-09 16:26:28.378889: Step 290: Cross entropy = 0.583013
2017-09-09 16:26:28.390262: Step 290: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:28.508627: Step 300: Train accuracy = 66.7%
2017-09-09 16:26:28.508670: Step 300: Cross entropy = 0.573335
2017-09-09 16:26:28.520155: Step 300: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:28.637557: Step 310: Train accuracy = 80.0%
2017-09-09 16:26:28.637598: Step 310: Cross entropy = 0.492504
2017-09-09 16:26:28.649032: Step 310: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:28.770088: Step 320: Train accuracy = 86.7%
2017-09-09 16:26:28.770129: Step 320: Cross entropy = 0.444201
2017-09-09 16:26:28.781680: Step 320: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:28.899976: Step 330: Train accuracy = 60.0%
2017-09-09 16:26:28.900019: Step 330: Cross entropy = 0.636856
2017-09-09 16:26:28.911438: Step 330: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:29.029704: Step 340: Train accuracy = 86.7%
2017-09-09 16:26:29.029746: Step 340: Cross entropy = 0.442068
2017-09-09 16:26:29.041351: Step 340: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:29.159450: Step 350: Train accuracy = 86.7%
2017-09-09 16:26:29.159491: Step 350: Cross entropy = 0.419123
2017-09-09 16:26:29.171113: Step 350: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:29.289170: Step 360: Train accuracy = 80.0%
2017-09-09 16:26:29.289212: Step 360: Cross entropy = 0.440261
2017-09-09 16:26:29.300641: Step 360: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:29.418111: Step 370: Train accuracy = 66.7%
2017-09-09 16:26:29.418154: Step 370: Cross entropy = 0.537198
2017-09-09 16:26:29.429563: Step 370: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:29.547782: Step 380: Train accuracy = 63.3%
2017-09-09 16:26:29.547823: Step 380: Cross entropy = 0.606099
2017-09-09 16:26:29.559285: Step 380: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:29.677531: Step 390: Train accuracy = 73.3%
2017-09-09 16:26:29.677572: Step 390: Cross entropy = 0.471767
2017-09-09 16:26:29.689097: Step 390: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:29.807420: Step 400: Train accuracy = 86.7%
2017-09-09 16:26:29.807461: Step 400: Cross entropy = 0.460176
2017-09-09 16:26:29.819012: Step 400: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:29.937846: Step 410: Train accuracy = 76.7%
2017-09-09 16:26:29.937885: Step 410: Cross entropy = 0.480160
2017-09-09 16:26:29.949341: Step 410: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:30.067236: Step 420: Train accuracy = 76.7%
2017-09-09 16:26:30.067279: Step 420: Cross entropy = 0.460587
2017-09-09 16:26:30.079398: Step 420: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:30.201588: Step 430: Train accuracy = 76.7%
2017-09-09 16:26:30.201632: Step 430: Cross entropy = 0.462227
2017-09-09 16:26:30.213016: Step 430: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:30.331723: Step 440: Train accuracy = 80.0%
2017-09-09 16:26:30.331765: Step 440: Cross entropy = 0.461432
2017-09-09 16:26:30.343321: Step 440: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:30.461565: Step 450: Train accuracy = 73.3%
2017-09-09 16:26:30.461606: Step 450: Cross entropy = 0.563950
2017-09-09 16:26:30.473123: Step 450: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:30.591360: Step 460: Train accuracy = 66.7%
2017-09-09 16:26:30.591403: Step 460: Cross entropy = 0.626901
2017-09-09 16:26:30.602945: Step 460: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:30.721163: Step 470: Train accuracy = 76.7%
2017-09-09 16:26:30.721203: Step 470: Cross entropy = 0.468814
2017-09-09 16:26:30.732667: Step 470: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:30.851695: Step 480: Train accuracy = 73.3%
2017-09-09 16:26:30.851744: Step 480: Cross entropy = 0.534881
2017-09-09 16:26:30.863200: Step 480: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:30.981387: Step 490: Train accuracy = 76.7%
2017-09-09 16:26:30.981432: Step 490: Cross entropy = 0.497966
2017-09-09 16:26:30.992872: Step 490: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:31.111187: Step 500: Train accuracy = 86.7%
2017-09-09 16:26:31.111229: Step 500: Cross entropy = 0.418084
2017-09-09 16:26:31.122786: Step 500: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:31.241703: Step 510: Train accuracy = 80.0%
2017-09-09 16:26:31.241748: Step 510: Cross entropy = 0.454936
2017-09-09 16:26:31.253743: Step 510: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:31.373871: Step 520: Train accuracy = 70.0%
2017-09-09 16:26:31.373912: Step 520: Cross entropy = 0.500673
2017-09-09 16:26:31.385492: Step 520: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:31.505167: Step 530: Train accuracy = 60.0%
2017-09-09 16:26:31.505208: Step 530: Cross entropy = 0.637683
2017-09-09 16:26:31.516682: Step 530: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:31.635639: Step 540: Train accuracy = 76.7%
2017-09-09 16:26:31.635680: Step 540: Cross entropy = 0.489649
2017-09-09 16:26:31.647096: Step 540: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:31.765069: Step 550: Train accuracy = 80.0%
2017-09-09 16:26:31.765111: Step 550: Cross entropy = 0.463477
2017-09-09 16:26:31.776598: Step 550: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:31.894716: Step 560: Train accuracy = 80.0%
2017-09-09 16:26:31.894757: Step 560: Cross entropy = 0.426713
2017-09-09 16:26:31.906291: Step 560: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:32.023977: Step 570: Train accuracy = 70.0%
2017-09-09 16:26:32.024021: Step 570: Cross entropy = 0.556764
2017-09-09 16:26:32.035412: Step 570: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:32.152960: Step 580: Train accuracy = 83.3%
2017-09-09 16:26:32.153006: Step 580: Cross entropy = 0.460265
2017-09-09 16:26:32.164527: Step 580: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:32.282086: Step 590: Train accuracy = 73.3%
2017-09-09 16:26:32.282129: Step 590: Cross entropy = 0.439191
2017-09-09 16:26:32.293615: Step 590: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:32.411642: Step 600: Train accuracy = 66.7%
2017-09-09 16:26:32.411684: Step 600: Cross entropy = 0.582294
2017-09-09 16:26:32.423112: Step 600: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:32.541837: Step 610: Train accuracy = 83.3%
2017-09-09 16:26:32.541877: Step 610: Cross entropy = 0.454508
2017-09-09 16:26:32.553224: Step 610: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:32.672207: Step 620: Train accuracy = 86.7%
2017-09-09 16:26:32.672250: Step 620: Cross entropy = 0.455682
2017-09-09 16:26:32.683809: Step 620: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:32.801788: Step 630: Train accuracy = 80.0%
2017-09-09 16:26:32.801830: Step 630: Cross entropy = 0.428833
2017-09-09 16:26:32.813379: Step 630: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:32.931589: Step 640: Train accuracy = 80.0%
2017-09-09 16:26:32.931629: Step 640: Cross entropy = 0.471932
2017-09-09 16:26:32.943013: Step 640: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:33.061596: Step 650: Train accuracy = 73.3%
2017-09-09 16:26:33.061637: Step 650: Cross entropy = 0.463982
2017-09-09 16:26:33.073225: Step 650: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:33.191018: Step 660: Train accuracy = 86.7%
2017-09-09 16:26:33.191060: Step 660: Cross entropy = 0.434184
2017-09-09 16:26:33.202686: Step 660: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:33.321877: Step 670: Train accuracy = 80.0%
2017-09-09 16:26:33.321920: Step 670: Cross entropy = 0.476752
2017-09-09 16:26:33.333383: Step 670: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:33.452160: Step 680: Train accuracy = 83.3%
2017-09-09 16:26:33.452202: Step 680: Cross entropy = 0.420177
2017-09-09 16:26:33.464675: Step 680: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:33.582358: Step 690: Train accuracy = 73.3%
2017-09-09 16:26:33.582399: Step 690: Cross entropy = 0.510914
2017-09-09 16:26:33.593865: Step 690: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:33.712805: Step 700: Train accuracy = 90.0%
2017-09-09 16:26:33.712847: Step 700: Cross entropy = 0.428098
2017-09-09 16:26:33.724396: Step 700: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:33.842288: Step 710: Train accuracy = 80.0%
2017-09-09 16:26:33.842331: Step 710: Cross entropy = 0.424307
2017-09-09 16:26:33.853860: Step 710: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:33.972670: Step 720: Train accuracy = 76.7%
2017-09-09 16:26:33.972711: Step 720: Cross entropy = 0.512309
2017-09-09 16:26:33.984908: Step 720: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:34.104108: Step 730: Train accuracy = 63.3%
2017-09-09 16:26:34.104148: Step 730: Cross entropy = 0.546445
2017-09-09 16:26:34.115725: Step 730: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:34.234769: Step 740: Train accuracy = 70.0%
2017-09-09 16:26:34.234809: Step 740: Cross entropy = 0.491757
2017-09-09 16:26:34.246296: Step 740: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:34.364724: Step 750: Train accuracy = 76.7%
2017-09-09 16:26:34.364766: Step 750: Cross entropy = 0.481914
2017-09-09 16:26:34.376183: Step 750: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:34.494192: Step 760: Train accuracy = 80.0%
2017-09-09 16:26:34.494234: Step 760: Cross entropy = 0.477952
2017-09-09 16:26:34.505882: Step 760: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:34.624232: Step 770: Train accuracy = 76.7%
2017-09-09 16:26:34.624274: Step 770: Cross entropy = 0.527315
2017-09-09 16:26:34.635661: Step 770: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:34.753928: Step 780: Train accuracy = 76.7%
2017-09-09 16:26:34.753970: Step 780: Cross entropy = 0.566049
2017-09-09 16:26:34.765447: Step 780: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:34.884527: Step 790: Train accuracy = 73.3%
2017-09-09 16:26:34.884569: Step 790: Cross entropy = 0.531486
2017-09-09 16:26:34.896094: Step 790: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:35.014307: Step 800: Train accuracy = 70.0%
2017-09-09 16:26:35.014349: Step 800: Cross entropy = 0.584919
2017-09-09 16:26:35.025875: Step 800: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:35.145325: Step 810: Train accuracy = 86.7%
2017-09-09 16:26:35.145366: Step 810: Cross entropy = 0.415519
2017-09-09 16:26:35.156837: Step 810: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:35.275338: Step 820: Train accuracy = 76.7%
2017-09-09 16:26:35.275380: Step 820: Cross entropy = 0.472042
2017-09-09 16:26:35.286954: Step 820: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:35.404902: Step 830: Train accuracy = 53.3%
2017-09-09 16:26:35.404943: Step 830: Cross entropy = 0.609977
2017-09-09 16:26:35.416319: Step 830: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:35.534336: Step 840: Train accuracy = 73.3%
2017-09-09 16:26:35.534379: Step 840: Cross entropy = 0.465481
2017-09-09 16:26:35.545946: Step 840: Validation accuracy = 93.3% (N=30)
2017-09-09 16:26:35.664469: Step 850: Train accuracy = 80.0%
2017-09-09 16:26:35.664512: Step 850: Cross entropy = 0.410934
2017-09-09 16:26:35.676021: Step 850: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:35.794494: Step 860: Train accuracy = 86.7%
2017-09-09 16:26:35.794536: Step 860: Cross entropy = 0.433545
2017-09-09 16:26:35.807078: Step 860: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:35.925848: Step 870: Train accuracy = 63.3%
2017-09-09 16:26:35.925889: Step 870: Cross entropy = 0.553435
2017-09-09 16:26:35.937487: Step 870: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:36.056641: Step 880: Train accuracy = 76.7%
2017-09-09 16:26:36.056684: Step 880: Cross entropy = 0.456222
2017-09-09 16:26:36.068971: Step 880: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:36.187162: Step 890: Train accuracy = 66.7%
2017-09-09 16:26:36.187206: Step 890: Cross entropy = 0.485631
2017-09-09 16:26:36.198752: Step 890: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:36.318032: Step 900: Train accuracy = 86.7%
2017-09-09 16:26:36.318074: Step 900: Cross entropy = 0.360677
2017-09-09 16:26:36.329701: Step 900: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:36.447650: Step 910: Train accuracy = 73.3%
2017-09-09 16:26:36.447695: Step 910: Cross entropy = 0.490111
2017-09-09 16:26:36.459509: Step 910: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:36.578592: Step 920: Train accuracy = 83.3%
2017-09-09 16:26:36.578634: Step 920: Cross entropy = 0.478630
2017-09-09 16:26:36.590452: Step 920: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:36.708654: Step 930: Train accuracy = 86.7%
2017-09-09 16:26:36.708697: Step 930: Cross entropy = 0.461208
2017-09-09 16:26:36.720110: Step 930: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:36.838782: Step 940: Train accuracy = 80.0%
2017-09-09 16:26:36.838824: Step 940: Cross entropy = 0.507198
2017-09-09 16:26:36.850256: Step 940: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:36.969935: Step 950: Train accuracy = 70.0%
2017-09-09 16:26:36.969978: Step 950: Cross entropy = 0.515531
2017-09-09 16:26:36.981542: Step 950: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:37.100167: Step 960: Train accuracy = 63.3%
2017-09-09 16:26:37.100208: Step 960: Cross entropy = 0.498953
2017-09-09 16:26:37.112092: Step 960: Validation accuracy = 93.3% (N=30)
2017-09-09 16:26:37.230629: Step 970: Train accuracy = 73.3%
2017-09-09 16:26:37.230670: Step 970: Cross entropy = 0.496989
2017-09-09 16:26:37.242231: Step 970: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:37.360892: Step 980: Train accuracy = 83.3%
2017-09-09 16:26:37.360935: Step 980: Cross entropy = 0.433327
2017-09-09 16:26:37.373140: Step 980: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:37.492156: Step 990: Train accuracy = 80.0%
2017-09-09 16:26:37.492201: Step 990: Cross entropy = 0.504591
2017-09-09 16:26:37.503740: Step 990: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:37.622447: Step 1000: Train accuracy = 76.7%
2017-09-09 16:26:37.622489: Step 1000: Cross entropy = 0.490113
2017-09-09 16:26:37.633954: Step 1000: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:37.751772: Step 1010: Train accuracy = 76.7%
2017-09-09 16:26:37.751815: Step 1010: Cross entropy = 0.499986
2017-09-09 16:26:37.763236: Step 1010: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:37.881389: Step 1020: Train accuracy = 76.7%
2017-09-09 16:26:37.881430: Step 1020: Cross entropy = 0.515070
2017-09-09 16:26:37.893059: Step 1020: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:38.017560: Step 1030: Train accuracy = 76.7%
2017-09-09 16:26:38.017602: Step 1030: Cross entropy = 0.475897
2017-09-09 16:26:38.029140: Step 1030: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:38.147851: Step 1040: Train accuracy = 83.3%
2017-09-09 16:26:38.147893: Step 1040: Cross entropy = 0.422912
2017-09-09 16:26:38.159346: Step 1040: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:38.277522: Step 1050: Train accuracy = 76.7%
2017-09-09 16:26:38.277565: Step 1050: Cross entropy = 0.460053
2017-09-09 16:26:38.289186: Step 1050: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:38.406982: Step 1060: Train accuracy = 76.7%
2017-09-09 16:26:38.407022: Step 1060: Cross entropy = 0.489544
2017-09-09 16:26:38.418528: Step 1060: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:38.537879: Step 1070: Train accuracy = 73.3%
2017-09-09 16:26:38.537920: Step 1070: Cross entropy = 0.544157
2017-09-09 16:26:38.549553: Step 1070: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:38.668119: Step 1080: Train accuracy = 73.3%
2017-09-09 16:26:38.668161: Step 1080: Cross entropy = 0.495510
2017-09-09 16:26:38.679604: Step 1080: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:38.798111: Step 1090: Train accuracy = 70.0%
2017-09-09 16:26:38.798153: Step 1090: Cross entropy = 0.570975
2017-09-09 16:26:38.809833: Step 1090: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:38.927624: Step 1100: Train accuracy = 80.0%
2017-09-09 16:26:38.927666: Step 1100: Cross entropy = 0.433761
2017-09-09 16:26:38.939263: Step 1100: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:39.057558: Step 1110: Train accuracy = 80.0%
2017-09-09 16:26:39.057601: Step 1110: Cross entropy = 0.476203
2017-09-09 16:26:39.069231: Step 1110: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:39.188128: Step 1120: Train accuracy = 70.0%
2017-09-09 16:26:39.188172: Step 1120: Cross entropy = 0.493356
2017-09-09 16:26:39.199635: Step 1120: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:39.319175: Step 1130: Train accuracy = 76.7%
2017-09-09 16:26:39.319217: Step 1130: Cross entropy = 0.507395
2017-09-09 16:26:39.330670: Step 1130: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:39.448687: Step 1140: Train accuracy = 83.3%
2017-09-09 16:26:39.448728: Step 1140: Cross entropy = 0.399098
2017-09-09 16:26:39.460589: Step 1140: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:39.578354: Step 1150: Train accuracy = 76.7%
2017-09-09 16:26:39.578396: Step 1150: Cross entropy = 0.509878
2017-09-09 16:26:39.589902: Step 1150: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:39.707939: Step 1160: Train accuracy = 83.3%
2017-09-09 16:26:39.707981: Step 1160: Cross entropy = 0.429192
2017-09-09 16:26:39.719563: Step 1160: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:39.837164: Step 1170: Train accuracy = 73.3%
2017-09-09 16:26:39.837205: Step 1170: Cross entropy = 0.449782
2017-09-09 16:26:39.849645: Step 1170: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:39.969150: Step 1180: Train accuracy = 76.7%
2017-09-09 16:26:39.969191: Step 1180: Cross entropy = 0.512610
2017-09-09 16:26:39.980780: Step 1180: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:40.098672: Step 1190: Train accuracy = 83.3%
2017-09-09 16:26:40.098713: Step 1190: Cross entropy = 0.412403
2017-09-09 16:26:40.110176: Step 1190: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:40.228790: Step 1200: Train accuracy = 80.0%
2017-09-09 16:26:40.228833: Step 1200: Cross entropy = 0.496857
2017-09-09 16:26:40.240386: Step 1200: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:40.359634: Step 1210: Train accuracy = 80.0%
2017-09-09 16:26:40.359674: Step 1210: Cross entropy = 0.550133
2017-09-09 16:26:40.371272: Step 1210: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:40.494525: Step 1220: Train accuracy = 73.3%
2017-09-09 16:26:40.494569: Step 1220: Cross entropy = 0.536395
2017-09-09 16:26:40.506137: Step 1220: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:40.624647: Step 1230: Train accuracy = 73.3%
2017-09-09 16:26:40.624688: Step 1230: Cross entropy = 0.478930
2017-09-09 16:26:40.636175: Step 1230: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:40.755126: Step 1240: Train accuracy = 80.0%
2017-09-09 16:26:40.755169: Step 1240: Cross entropy = 0.428815
2017-09-09 16:26:40.766985: Step 1240: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:40.884705: Step 1250: Train accuracy = 60.0%
2017-09-09 16:26:40.884747: Step 1250: Cross entropy = 0.634753
2017-09-09 16:26:40.896248: Step 1250: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:41.013988: Step 1260: Train accuracy = 80.0%
2017-09-09 16:26:41.014029: Step 1260: Cross entropy = 0.439792
2017-09-09 16:26:41.025463: Step 1260: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:41.143958: Step 1270: Train accuracy = 83.3%
2017-09-09 16:26:41.144001: Step 1270: Cross entropy = 0.450133
2017-09-09 16:26:41.156429: Step 1270: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:41.280064: Step 1280: Train accuracy = 70.0%
2017-09-09 16:26:41.280105: Step 1280: Cross entropy = 0.587717
2017-09-09 16:26:41.291577: Step 1280: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:41.409523: Step 1290: Train accuracy = 76.7%
2017-09-09 16:26:41.409565: Step 1290: Cross entropy = 0.417555
2017-09-09 16:26:41.421074: Step 1290: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:41.539443: Step 1300: Train accuracy = 83.3%
2017-09-09 16:26:41.539486: Step 1300: Cross entropy = 0.405848
2017-09-09 16:26:41.551015: Step 1300: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:41.670357: Step 1310: Train accuracy = 60.0%
2017-09-09 16:26:41.670398: Step 1310: Cross entropy = 0.578272
2017-09-09 16:26:41.681974: Step 1310: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:41.802810: Step 1320: Train accuracy = 73.3%
2017-09-09 16:26:41.802853: Step 1320: Cross entropy = 0.443420
2017-09-09 16:26:41.814381: Step 1320: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:41.934338: Step 1330: Train accuracy = 76.7%
2017-09-09 16:26:41.934380: Step 1330: Cross entropy = 0.515278
2017-09-09 16:26:41.945882: Step 1330: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:42.067949: Step 1340: Train accuracy = 83.3%
2017-09-09 16:26:42.067991: Step 1340: Cross entropy = 0.405957
2017-09-09 16:26:42.079561: Step 1340: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:42.201455: Step 1350: Train accuracy = 86.7%
2017-09-09 16:26:42.201496: Step 1350: Cross entropy = 0.391114
2017-09-09 16:26:42.213410: Step 1350: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:42.335778: Step 1360: Train accuracy = 86.7%
2017-09-09 16:26:42.335820: Step 1360: Cross entropy = 0.460106
2017-09-09 16:26:42.347438: Step 1360: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:42.466705: Step 1370: Train accuracy = 70.0%
2017-09-09 16:26:42.466747: Step 1370: Cross entropy = 0.523801
2017-09-09 16:26:42.478207: Step 1370: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:42.599119: Step 1380: Train accuracy = 73.3%
2017-09-09 16:26:42.599163: Step 1380: Cross entropy = 0.481830
2017-09-09 16:26:42.610673: Step 1380: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:42.729800: Step 1390: Train accuracy = 76.7%
2017-09-09 16:26:42.729843: Step 1390: Cross entropy = 0.493771
2017-09-09 16:26:42.741738: Step 1390: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:42.868345: Step 1400: Train accuracy = 80.0%
2017-09-09 16:26:42.868389: Step 1400: Cross entropy = 0.492628
2017-09-09 16:26:42.880169: Step 1400: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:43.001278: Step 1410: Train accuracy = 66.7%
2017-09-09 16:26:43.001319: Step 1410: Cross entropy = 0.467376
2017-09-09 16:26:43.013071: Step 1410: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:43.134661: Step 1420: Train accuracy = 83.3%
2017-09-09 16:26:43.134704: Step 1420: Cross entropy = 0.422643
2017-09-09 16:26:43.146403: Step 1420: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:43.268548: Step 1430: Train accuracy = 86.7%
2017-09-09 16:26:43.268592: Step 1430: Cross entropy = 0.413714
2017-09-09 16:26:43.280202: Step 1430: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:43.400250: Step 1440: Train accuracy = 80.0%
2017-09-09 16:26:43.400293: Step 1440: Cross entropy = 0.471109
2017-09-09 16:26:43.411835: Step 1440: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:43.530985: Step 1450: Train accuracy = 80.0%
2017-09-09 16:26:43.531028: Step 1450: Cross entropy = 0.467988
2017-09-09 16:26:43.542566: Step 1450: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:43.662421: Step 1460: Train accuracy = 80.0%
2017-09-09 16:26:43.662465: Step 1460: Cross entropy = 0.504426
2017-09-09 16:26:43.674040: Step 1460: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:43.792229: Step 1470: Train accuracy = 80.0%
2017-09-09 16:26:43.792272: Step 1470: Cross entropy = 0.482122
2017-09-09 16:26:43.803808: Step 1470: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:43.927089: Step 1480: Train accuracy = 70.0%
2017-09-09 16:26:43.927131: Step 1480: Cross entropy = 0.493073
2017-09-09 16:26:43.938731: Step 1480: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:44.064125: Step 1490: Train accuracy = 73.3%
2017-09-09 16:26:44.064168: Step 1490: Cross entropy = 0.563185
2017-09-09 16:26:44.075887: Step 1490: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:44.213618: Step 1500: Train accuracy = 80.0%
2017-09-09 16:26:44.213661: Step 1500: Cross entropy = 0.506072
2017-09-09 16:26:44.225984: Step 1500: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:44.354155: Step 1510: Train accuracy = 83.3%
2017-09-09 16:26:44.354198: Step 1510: Cross entropy = 0.458797
2017-09-09 16:26:44.367181: Step 1510: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:44.493282: Step 1520: Train accuracy = 80.0%
2017-09-09 16:26:44.493324: Step 1520: Cross entropy = 0.484274
2017-09-09 16:26:44.505069: Step 1520: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:44.626556: Step 1530: Train accuracy = 83.3%
2017-09-09 16:26:44.626597: Step 1530: Cross entropy = 0.410692
2017-09-09 16:26:44.638754: Step 1530: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:44.761011: Step 1540: Train accuracy = 90.0%
2017-09-09 16:26:44.761052: Step 1540: Cross entropy = 0.388571
2017-09-09 16:26:44.772899: Step 1540: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:44.894256: Step 1550: Train accuracy = 86.7%
2017-09-09 16:26:44.894298: Step 1550: Cross entropy = 0.373561
2017-09-09 16:26:44.906019: Step 1550: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:45.029035: Step 1560: Train accuracy = 90.0%
2017-09-09 16:26:45.029079: Step 1560: Cross entropy = 0.386103
2017-09-09 16:26:45.040792: Step 1560: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:45.162039: Step 1570: Train accuracy = 80.0%
2017-09-09 16:26:45.162082: Step 1570: Cross entropy = 0.412631
2017-09-09 16:26:45.173820: Step 1570: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:45.295093: Step 1580: Train accuracy = 80.0%
2017-09-09 16:26:45.295136: Step 1580: Cross entropy = 0.410892
2017-09-09 16:26:45.306801: Step 1580: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:45.425664: Step 1590: Train accuracy = 90.0%
2017-09-09 16:26:45.425706: Step 1590: Cross entropy = 0.379144
2017-09-09 16:26:45.437184: Step 1590: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:45.557922: Step 1600: Train accuracy = 63.3%
2017-09-09 16:26:45.557966: Step 1600: Cross entropy = 0.638779
2017-09-09 16:26:45.569771: Step 1600: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:45.691703: Step 1610: Train accuracy = 90.0%
2017-09-09 16:26:45.691754: Step 1610: Cross entropy = 0.336658
2017-09-09 16:26:45.703270: Step 1610: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:45.826779: Step 1620: Train accuracy = 70.0%
2017-09-09 16:26:45.826820: Step 1620: Cross entropy = 0.508588
2017-09-09 16:26:45.838353: Step 1620: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:45.958062: Step 1630: Train accuracy = 73.3%
2017-09-09 16:26:45.958104: Step 1630: Cross entropy = 0.527622
2017-09-09 16:26:45.969721: Step 1630: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:46.087563: Step 1640: Train accuracy = 90.0%
2017-09-09 16:26:46.087604: Step 1640: Cross entropy = 0.329633
2017-09-09 16:26:46.099117: Step 1640: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:46.217193: Step 1650: Train accuracy = 66.7%
2017-09-09 16:26:46.217235: Step 1650: Cross entropy = 0.599353
2017-09-09 16:26:46.229067: Step 1650: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:46.349326: Step 1660: Train accuracy = 93.3%
2017-09-09 16:26:46.349370: Step 1660: Cross entropy = 0.384008
2017-09-09 16:26:46.361116: Step 1660: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:46.483033: Step 1670: Train accuracy = 56.7%
2017-09-09 16:26:46.483088: Step 1670: Cross entropy = 0.709601
2017-09-09 16:26:46.494908: Step 1670: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:46.616856: Step 1680: Train accuracy = 90.0%
2017-09-09 16:26:46.616898: Step 1680: Cross entropy = 0.391978
2017-09-09 16:26:46.628542: Step 1680: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:46.748218: Step 1690: Train accuracy = 80.0%
2017-09-09 16:26:46.748259: Step 1690: Cross entropy = 0.408767
2017-09-09 16:26:46.760172: Step 1690: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:46.882476: Step 1700: Train accuracy = 76.7%
2017-09-09 16:26:46.882519: Step 1700: Cross entropy = 0.454543
2017-09-09 16:26:46.894481: Step 1700: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:47.017672: Step 1710: Train accuracy = 90.0%
2017-09-09 16:26:47.017714: Step 1710: Cross entropy = 0.363475
2017-09-09 16:26:47.029491: Step 1710: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:47.151522: Step 1720: Train accuracy = 76.7%
2017-09-09 16:26:47.151566: Step 1720: Cross entropy = 0.526508
2017-09-09 16:26:47.163491: Step 1720: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:47.286422: Step 1730: Train accuracy = 86.7%
2017-09-09 16:26:47.286465: Step 1730: Cross entropy = 0.399168
2017-09-09 16:26:47.298210: Step 1730: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:47.420687: Step 1740: Train accuracy = 73.3%
2017-09-09 16:26:47.420732: Step 1740: Cross entropy = 0.485571
2017-09-09 16:26:47.433158: Step 1740: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:47.552349: Step 1750: Train accuracy = 76.7%
2017-09-09 16:26:47.552392: Step 1750: Cross entropy = 0.392643
2017-09-09 16:26:47.564052: Step 1750: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:47.685439: Step 1760: Train accuracy = 73.3%
2017-09-09 16:26:47.685481: Step 1760: Cross entropy = 0.536980
2017-09-09 16:26:47.697241: Step 1760: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:47.816598: Step 1770: Train accuracy = 83.3%
2017-09-09 16:26:47.816644: Step 1770: Cross entropy = 0.483710
2017-09-09 16:26:47.828138: Step 1770: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:47.945715: Step 1780: Train accuracy = 70.0%
2017-09-09 16:26:47.945757: Step 1780: Cross entropy = 0.530328
2017-09-09 16:26:47.958136: Step 1780: Validation accuracy = 53.3% (N=30)
2017-09-09 16:26:48.077136: Step 1790: Train accuracy = 76.7%
2017-09-09 16:26:48.077177: Step 1790: Cross entropy = 0.540950
2017-09-09 16:26:48.088790: Step 1790: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:48.207435: Step 1800: Train accuracy = 73.3%
2017-09-09 16:26:48.207482: Step 1800: Cross entropy = 0.539904
2017-09-09 16:26:48.218973: Step 1800: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:48.339028: Step 1810: Train accuracy = 76.7%
2017-09-09 16:26:48.339070: Step 1810: Cross entropy = 0.536756
2017-09-09 16:26:48.350660: Step 1810: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:48.470882: Step 1820: Train accuracy = 80.0%
2017-09-09 16:26:48.470923: Step 1820: Cross entropy = 0.474241
2017-09-09 16:26:48.482538: Step 1820: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:48.601300: Step 1830: Train accuracy = 83.3%
2017-09-09 16:26:48.601342: Step 1830: Cross entropy = 0.408922
2017-09-09 16:26:48.612873: Step 1830: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:48.731120: Step 1840: Train accuracy = 83.3%
2017-09-09 16:26:48.731164: Step 1840: Cross entropy = 0.407886
2017-09-09 16:26:48.742602: Step 1840: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:48.861925: Step 1850: Train accuracy = 83.3%
2017-09-09 16:26:48.861967: Step 1850: Cross entropy = 0.406712
2017-09-09 16:26:48.873512: Step 1850: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:48.994581: Step 1860: Train accuracy = 70.0%
2017-09-09 16:26:48.994624: Step 1860: Cross entropy = 0.568639
2017-09-09 16:26:49.006170: Step 1860: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:49.129047: Step 1870: Train accuracy = 76.7%
2017-09-09 16:26:49.129091: Step 1870: Cross entropy = 0.403455
2017-09-09 16:26:49.140876: Step 1870: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:49.265392: Step 1880: Train accuracy = 93.3%
2017-09-09 16:26:49.265435: Step 1880: Cross entropy = 0.366471
2017-09-09 16:26:49.277213: Step 1880: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:49.400375: Step 1890: Train accuracy = 86.7%
2017-09-09 16:26:49.400419: Step 1890: Cross entropy = 0.420193
2017-09-09 16:26:49.412323: Step 1890: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:49.534388: Step 1900: Train accuracy = 86.7%
2017-09-09 16:26:49.534437: Step 1900: Cross entropy = 0.360307
2017-09-09 16:26:49.546328: Step 1900: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:49.669470: Step 1910: Train accuracy = 83.3%
2017-09-09 16:26:49.669512: Step 1910: Cross entropy = 0.487052
2017-09-09 16:26:49.681240: Step 1910: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:49.803171: Step 1920: Train accuracy = 60.0%
2017-09-09 16:26:49.803213: Step 1920: Cross entropy = 0.652123
2017-09-09 16:26:49.814933: Step 1920: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:49.937086: Step 1930: Train accuracy = 73.3%
2017-09-09 16:26:49.937130: Step 1930: Cross entropy = 0.528464
2017-09-09 16:26:49.948758: Step 1930: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:50.066974: Step 1940: Train accuracy = 76.7%
2017-09-09 16:26:50.067017: Step 1940: Cross entropy = 0.445566
2017-09-09 16:26:50.078455: Step 1940: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:50.197067: Step 1950: Train accuracy = 83.3%
2017-09-09 16:26:50.197110: Step 1950: Cross entropy = 0.353406
2017-09-09 16:26:50.208651: Step 1950: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:50.328427: Step 1960: Train accuracy = 76.7%
2017-09-09 16:26:50.328469: Step 1960: Cross entropy = 0.490217
2017-09-09 16:26:50.340011: Step 1960: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:50.459258: Step 1970: Train accuracy = 76.7%
2017-09-09 16:26:50.459299: Step 1970: Cross entropy = 0.470098
2017-09-09 16:26:50.470693: Step 1970: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:50.590651: Step 1980: Train accuracy = 76.7%
2017-09-09 16:26:50.590693: Step 1980: Cross entropy = 0.428328
2017-09-09 16:26:50.602218: Step 1980: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:50.721494: Step 1990: Train accuracy = 80.0%
2017-09-09 16:26:50.721539: Step 1990: Cross entropy = 0.374904
2017-09-09 16:26:50.733115: Step 1990: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:50.853619: Step 2000: Train accuracy = 86.7%
2017-09-09 16:26:50.853664: Step 2000: Cross entropy = 0.362862
2017-09-09 16:26:50.865108: Step 2000: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:50.984856: Step 2010: Train accuracy = 66.7%
2017-09-09 16:26:50.984898: Step 2010: Cross entropy = 0.566798
2017-09-09 16:26:50.996250: Step 2010: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:51.115979: Step 2020: Train accuracy = 80.0%
2017-09-09 16:26:51.116021: Step 2020: Cross entropy = 0.371098
2017-09-09 16:26:51.127759: Step 2020: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:51.246374: Step 2030: Train accuracy = 80.0%
2017-09-09 16:26:51.246416: Step 2030: Cross entropy = 0.387635
2017-09-09 16:26:51.257902: Step 2030: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:51.377216: Step 2040: Train accuracy = 80.0%
2017-09-09 16:26:51.377257: Step 2040: Cross entropy = 0.423347
2017-09-09 16:26:51.388657: Step 2040: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:51.506805: Step 2050: Train accuracy = 80.0%
2017-09-09 16:26:51.506846: Step 2050: Cross entropy = 0.490842
2017-09-09 16:26:51.518253: Step 2050: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:51.637721: Step 2060: Train accuracy = 76.7%
2017-09-09 16:26:51.637762: Step 2060: Cross entropy = 0.444490
2017-09-09 16:26:51.649997: Step 2060: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:51.769905: Step 2070: Train accuracy = 93.3%
2017-09-09 16:26:51.769946: Step 2070: Cross entropy = 0.332800
2017-09-09 16:26:51.781560: Step 2070: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:51.903371: Step 2080: Train accuracy = 83.3%
2017-09-09 16:26:51.903412: Step 2080: Cross entropy = 0.379852
2017-09-09 16:26:51.915114: Step 2080: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:52.036384: Step 2090: Train accuracy = 83.3%
2017-09-09 16:26:52.036427: Step 2090: Cross entropy = 0.374388
2017-09-09 16:26:52.048006: Step 2090: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:52.170049: Step 2100: Train accuracy = 76.7%
2017-09-09 16:26:52.170093: Step 2100: Cross entropy = 0.442556
2017-09-09 16:26:52.181920: Step 2100: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:52.303113: Step 2110: Train accuracy = 93.3%
2017-09-09 16:26:52.303157: Step 2110: Cross entropy = 0.297073
2017-09-09 16:26:52.314782: Step 2110: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:52.436360: Step 2120: Train accuracy = 90.0%
2017-09-09 16:26:52.436403: Step 2120: Cross entropy = 0.442199
2017-09-09 16:26:52.448051: Step 2120: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:52.569333: Step 2130: Train accuracy = 86.7%
2017-09-09 16:26:52.569376: Step 2130: Cross entropy = 0.459332
2017-09-09 16:26:52.580936: Step 2130: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:52.702808: Step 2140: Train accuracy = 63.3%
2017-09-09 16:26:52.702849: Step 2140: Cross entropy = 0.610277
2017-09-09 16:26:52.714806: Step 2140: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:52.835128: Step 2150: Train accuracy = 83.3%
2017-09-09 16:26:52.835169: Step 2150: Cross entropy = 0.387783
2017-09-09 16:26:52.846586: Step 2150: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:52.965180: Step 2160: Train accuracy = 83.3%
2017-09-09 16:26:52.965223: Step 2160: Cross entropy = 0.429226
2017-09-09 16:26:52.976799: Step 2160: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:53.096896: Step 2170: Train accuracy = 76.7%
2017-09-09 16:26:53.096937: Step 2170: Cross entropy = 0.420783
2017-09-09 16:26:53.108615: Step 2170: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:53.228489: Step 2180: Train accuracy = 66.7%
2017-09-09 16:26:53.228530: Step 2180: Cross entropy = 0.491212
2017-09-09 16:26:53.240136: Step 2180: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:53.358722: Step 2190: Train accuracy = 66.7%
2017-09-09 16:26:53.358763: Step 2190: Cross entropy = 0.626561
2017-09-09 16:26:53.370452: Step 2190: Validation accuracy = 93.3% (N=30)
2017-09-09 16:26:53.489498: Step 2200: Train accuracy = 76.7%
2017-09-09 16:26:53.489540: Step 2200: Cross entropy = 0.487547
2017-09-09 16:26:53.501081: Step 2200: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:53.619325: Step 2210: Train accuracy = 80.0%
2017-09-09 16:26:53.619368: Step 2210: Cross entropy = 0.476456
2017-09-09 16:26:53.630819: Step 2210: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:53.748585: Step 2220: Train accuracy = 83.3%
2017-09-09 16:26:53.748627: Step 2220: Cross entropy = 0.387242
2017-09-09 16:26:53.759999: Step 2220: Validation accuracy = 56.7% (N=30)
2017-09-09 16:26:53.880819: Step 2230: Train accuracy = 93.3%
2017-09-09 16:26:53.880860: Step 2230: Cross entropy = 0.354563
2017-09-09 16:26:53.892321: Step 2230: Validation accuracy = 66.7% (N=30)
2017-09-09 16:26:54.011150: Step 2240: Train accuracy = 76.7%
2017-09-09 16:26:54.011192: Step 2240: Cross entropy = 0.524362
2017-09-09 16:26:54.022744: Step 2240: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:54.141391: Step 2250: Train accuracy = 83.3%
2017-09-09 16:26:54.141433: Step 2250: Cross entropy = 0.408758
2017-09-09 16:26:54.153013: Step 2250: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:54.271652: Step 2260: Train accuracy = 66.7%
2017-09-09 16:26:54.271694: Step 2260: Cross entropy = 0.570952
2017-09-09 16:26:54.283236: Step 2260: Validation accuracy = 60.0% (N=30)
2017-09-09 16:26:54.403800: Step 2270: Train accuracy = 70.0%
2017-09-09 16:26:54.403841: Step 2270: Cross entropy = 0.530580
2017-09-09 16:26:54.416415: Step 2270: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:54.538300: Step 2280: Train accuracy = 86.7%
2017-09-09 16:26:54.538343: Step 2280: Cross entropy = 0.380104
2017-09-09 16:26:54.550076: Step 2280: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:54.673173: Step 2290: Train accuracy = 83.3%
2017-09-09 16:26:54.673215: Step 2290: Cross entropy = 0.319364
2017-09-09 16:26:54.684822: Step 2290: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:54.806368: Step 2300: Train accuracy = 76.7%
2017-09-09 16:26:54.806410: Step 2300: Cross entropy = 0.491547
2017-09-09 16:26:54.818110: Step 2300: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:54.939226: Step 2310: Train accuracy = 66.7%
2017-09-09 16:26:54.939268: Step 2310: Cross entropy = 0.519702
2017-09-09 16:26:54.951102: Step 2310: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:55.089475: Step 2320: Train accuracy = 80.0%
2017-09-09 16:26:55.089519: Step 2320: Cross entropy = 0.444519
2017-09-09 16:26:55.101286: Step 2320: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:55.222844: Step 2330: Train accuracy = 80.0%
2017-09-09 16:26:55.222888: Step 2330: Cross entropy = 0.478530
2017-09-09 16:26:55.234670: Step 2330: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:55.358981: Step 2340: Train accuracy = 73.3%
2017-09-09 16:26:55.359024: Step 2340: Cross entropy = 0.483399
2017-09-09 16:26:55.370948: Step 2340: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:55.490751: Step 2350: Train accuracy = 80.0%
2017-09-09 16:26:55.490792: Step 2350: Cross entropy = 0.476355
2017-09-09 16:26:55.502363: Step 2350: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:55.621203: Step 2360: Train accuracy = 93.3%
2017-09-09 16:26:55.621245: Step 2360: Cross entropy = 0.280858
2017-09-09 16:26:55.632700: Step 2360: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:55.752658: Step 2370: Train accuracy = 76.7%
2017-09-09 16:26:55.752699: Step 2370: Cross entropy = 0.472890
2017-09-09 16:26:55.764176: Step 2370: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:55.885849: Step 2380: Train accuracy = 83.3%
2017-09-09 16:26:55.885893: Step 2380: Cross entropy = 0.410608
2017-09-09 16:26:55.898080: Step 2380: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:56.016970: Step 2390: Train accuracy = 93.3%
2017-09-09 16:26:56.017012: Step 2390: Cross entropy = 0.351915
2017-09-09 16:26:56.028659: Step 2390: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:56.148599: Step 2400: Train accuracy = 86.7%
2017-09-09 16:26:56.148641: Step 2400: Cross entropy = 0.353593
2017-09-09 16:26:56.160099: Step 2400: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:56.281711: Step 2410: Train accuracy = 86.7%
2017-09-09 16:26:56.281752: Step 2410: Cross entropy = 0.405933
2017-09-09 16:26:56.293339: Step 2410: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:56.414016: Step 2420: Train accuracy = 73.3%
2017-09-09 16:26:56.414057: Step 2420: Cross entropy = 0.475250
2017-09-09 16:26:56.425465: Step 2420: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:56.543431: Step 2430: Train accuracy = 83.3%
2017-09-09 16:26:56.543472: Step 2430: Cross entropy = 0.392055
2017-09-09 16:26:56.555018: Step 2430: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:56.673687: Step 2440: Train accuracy = 83.3%
2017-09-09 16:26:56.673729: Step 2440: Cross entropy = 0.425877
2017-09-09 16:26:56.685118: Step 2440: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:56.803779: Step 2450: Train accuracy = 83.3%
2017-09-09 16:26:56.803822: Step 2450: Cross entropy = 0.447921
2017-09-09 16:26:56.815469: Step 2450: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:56.936995: Step 2460: Train accuracy = 90.0%
2017-09-09 16:26:56.937038: Step 2460: Cross entropy = 0.364484
2017-09-09 16:26:56.948799: Step 2460: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:57.069959: Step 2470: Train accuracy = 80.0%
2017-09-09 16:26:57.070000: Step 2470: Cross entropy = 0.494192
2017-09-09 16:26:57.082676: Step 2470: Validation accuracy = 93.3% (N=30)
2017-09-09 16:26:57.204693: Step 2480: Train accuracy = 80.0%
2017-09-09 16:26:57.204738: Step 2480: Cross entropy = 0.422003
2017-09-09 16:26:57.216732: Step 2480: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:57.337573: Step 2490: Train accuracy = 80.0%
2017-09-09 16:26:57.337617: Step 2490: Cross entropy = 0.482975
2017-09-09 16:26:57.349710: Step 2490: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:57.471615: Step 2500: Train accuracy = 80.0%
2017-09-09 16:26:57.471656: Step 2500: Cross entropy = 0.433664
2017-09-09 16:26:57.483743: Step 2500: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:57.604927: Step 2510: Train accuracy = 86.7%
2017-09-09 16:26:57.604968: Step 2510: Cross entropy = 0.326721
2017-09-09 16:26:57.616903: Step 2510: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:57.739532: Step 2520: Train accuracy = 93.3%
2017-09-09 16:26:57.739577: Step 2520: Cross entropy = 0.305844
2017-09-09 16:26:57.751157: Step 2520: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:57.872586: Step 2530: Train accuracy = 83.3%
2017-09-09 16:26:57.872629: Step 2530: Cross entropy = 0.374446
2017-09-09 16:26:57.884153: Step 2530: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:58.002282: Step 2540: Train accuracy = 86.7%
2017-09-09 16:26:58.002323: Step 2540: Cross entropy = 0.386164
2017-09-09 16:26:58.014065: Step 2540: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:58.132153: Step 2550: Train accuracy = 83.3%
2017-09-09 16:26:58.132195: Step 2550: Cross entropy = 0.318151
2017-09-09 16:26:58.143694: Step 2550: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:58.262130: Step 2560: Train accuracy = 86.7%
2017-09-09 16:26:58.262172: Step 2560: Cross entropy = 0.351709
2017-09-09 16:26:58.273693: Step 2560: Validation accuracy = 86.7% (N=30)
2017-09-09 16:26:58.391324: Step 2570: Train accuracy = 86.7%
2017-09-09 16:26:58.391366: Step 2570: Cross entropy = 0.296772
2017-09-09 16:26:58.402820: Step 2570: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:58.521024: Step 2580: Train accuracy = 86.7%
2017-09-09 16:26:58.521065: Step 2580: Cross entropy = 0.348841
2017-09-09 16:26:58.532592: Step 2580: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:58.650392: Step 2590: Train accuracy = 83.3%
2017-09-09 16:26:58.650435: Step 2590: Cross entropy = 0.370972
2017-09-09 16:26:58.661943: Step 2590: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:58.780431: Step 2600: Train accuracy = 76.7%
2017-09-09 16:26:58.780472: Step 2600: Cross entropy = 0.495526
2017-09-09 16:26:58.792067: Step 2600: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:58.909584: Step 2610: Train accuracy = 76.7%
2017-09-09 16:26:58.909626: Step 2610: Cross entropy = 0.462559
2017-09-09 16:26:58.920981: Step 2610: Validation accuracy = 76.7% (N=30)
2017-09-09 16:26:59.040053: Step 2620: Train accuracy = 76.7%
2017-09-09 16:26:59.040094: Step 2620: Cross entropy = 0.523050
2017-09-09 16:26:59.051567: Step 2620: Validation accuracy = 70.0% (N=30)
2017-09-09 16:26:59.169847: Step 2630: Train accuracy = 86.7%
2017-09-09 16:26:59.169892: Step 2630: Cross entropy = 0.351420
2017-09-09 16:26:59.181282: Step 2630: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:59.300856: Step 2640: Train accuracy = 80.0%
2017-09-09 16:26:59.300898: Step 2640: Cross entropy = 0.338965
2017-09-09 16:26:59.312523: Step 2640: Validation accuracy = 80.0% (N=30)
2017-09-09 16:26:59.433350: Step 2650: Train accuracy = 76.7%
2017-09-09 16:26:59.433391: Step 2650: Cross entropy = 0.507768
2017-09-09 16:26:59.444968: Step 2650: Validation accuracy = 90.0% (N=30)
2017-09-09 16:26:59.565935: Step 2660: Train accuracy = 83.3%
2017-09-09 16:26:59.565982: Step 2660: Cross entropy = 0.389046
2017-09-09 16:26:59.577559: Step 2660: Validation accuracy = 63.3% (N=30)
2017-09-09 16:26:59.699519: Step 2670: Train accuracy = 73.3%
2017-09-09 16:26:59.699564: Step 2670: Cross entropy = 0.500239
2017-09-09 16:26:59.712143: Step 2670: Validation accuracy = 73.3% (N=30)
2017-09-09 16:26:59.833909: Step 2680: Train accuracy = 80.0%
2017-09-09 16:26:59.833953: Step 2680: Cross entropy = 0.410539
2017-09-09 16:26:59.845563: Step 2680: Validation accuracy = 83.3% (N=30)
2017-09-09 16:26:59.966307: Step 2690: Train accuracy = 66.7%
2017-09-09 16:26:59.966349: Step 2690: Cross entropy = 0.582248
2017-09-09 16:26:59.978086: Step 2690: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:00.099210: Step 2700: Train accuracy = 86.7%
2017-09-09 16:27:00.099251: Step 2700: Cross entropy = 0.332246
2017-09-09 16:27:00.110897: Step 2700: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:00.231920: Step 2710: Train accuracy = 76.7%
2017-09-09 16:27:00.231966: Step 2710: Cross entropy = 0.404456
2017-09-09 16:27:00.243581: Step 2710: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:00.364593: Step 2720: Train accuracy = 76.7%
2017-09-09 16:27:00.364634: Step 2720: Cross entropy = 0.453767
2017-09-09 16:27:00.376104: Step 2720: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:00.494236: Step 2730: Train accuracy = 83.3%
2017-09-09 16:27:00.494278: Step 2730: Cross entropy = 0.380876
2017-09-09 16:27:00.505860: Step 2730: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:00.624743: Step 2740: Train accuracy = 80.0%
2017-09-09 16:27:00.624784: Step 2740: Cross entropy = 0.381388
2017-09-09 16:27:00.636281: Step 2740: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:00.754513: Step 2750: Train accuracy = 83.3%
2017-09-09 16:27:00.754559: Step 2750: Cross entropy = 0.448840
2017-09-09 16:27:00.765988: Step 2750: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:00.884468: Step 2760: Train accuracy = 83.3%
2017-09-09 16:27:00.884510: Step 2760: Cross entropy = 0.382744
2017-09-09 16:27:00.895956: Step 2760: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:01.014378: Step 2770: Train accuracy = 66.7%
2017-09-09 16:27:01.014419: Step 2770: Cross entropy = 0.604957
2017-09-09 16:27:01.026088: Step 2770: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:01.144736: Step 2780: Train accuracy = 80.0%
2017-09-09 16:27:01.144777: Step 2780: Cross entropy = 0.428042
2017-09-09 16:27:01.156175: Step 2780: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:01.274206: Step 2790: Train accuracy = 86.7%
2017-09-09 16:27:01.274250: Step 2790: Cross entropy = 0.386429
2017-09-09 16:27:01.285736: Step 2790: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:01.404635: Step 2800: Train accuracy = 76.7%
2017-09-09 16:27:01.404677: Step 2800: Cross entropy = 0.479230
2017-09-09 16:27:01.416302: Step 2800: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:01.533850: Step 2810: Train accuracy = 73.3%
2017-09-09 16:27:01.533892: Step 2810: Cross entropy = 0.460472
2017-09-09 16:27:01.545530: Step 2810: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:01.663931: Step 2820: Train accuracy = 83.3%
2017-09-09 16:27:01.663971: Step 2820: Cross entropy = 0.520382
2017-09-09 16:27:01.675601: Step 2820: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:01.795170: Step 2830: Train accuracy = 86.7%
2017-09-09 16:27:01.795211: Step 2830: Cross entropy = 0.418565
2017-09-09 16:27:01.806712: Step 2830: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:01.927978: Step 2840: Train accuracy = 96.7%
2017-09-09 16:27:01.928022: Step 2840: Cross entropy = 0.243654
2017-09-09 16:27:01.940075: Step 2840: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:02.061068: Step 2850: Train accuracy = 93.3%
2017-09-09 16:27:02.061113: Step 2850: Cross entropy = 0.300443
2017-09-09 16:27:02.072909: Step 2850: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:02.193627: Step 2860: Train accuracy = 90.0%
2017-09-09 16:27:02.193669: Step 2860: Cross entropy = 0.358909
2017-09-09 16:27:02.205282: Step 2860: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:02.326334: Step 2870: Train accuracy = 90.0%
2017-09-09 16:27:02.326375: Step 2870: Cross entropy = 0.349378
2017-09-09 16:27:02.337926: Step 2870: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:02.459095: Step 2880: Train accuracy = 73.3%
2017-09-09 16:27:02.459137: Step 2880: Cross entropy = 0.571198
2017-09-09 16:27:02.471100: Step 2880: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:02.594869: Step 2890: Train accuracy = 73.3%
2017-09-09 16:27:02.594911: Step 2890: Cross entropy = 0.481559
2017-09-09 16:27:02.606714: Step 2890: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:02.729683: Step 2900: Train accuracy = 90.0%
2017-09-09 16:27:02.729725: Step 2900: Cross entropy = 0.347378
2017-09-09 16:27:02.741621: Step 2900: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:02.866092: Step 2910: Train accuracy = 83.3%
2017-09-09 16:27:02.866135: Step 2910: Cross entropy = 0.459075
2017-09-09 16:27:02.878176: Step 2910: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:02.998224: Step 2920: Train accuracy = 90.0%
2017-09-09 16:27:02.998266: Step 2920: Cross entropy = 0.328280
2017-09-09 16:27:03.009704: Step 2920: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:03.128692: Step 2930: Train accuracy = 83.3%
2017-09-09 16:27:03.128735: Step 2930: Cross entropy = 0.461842
2017-09-09 16:27:03.140181: Step 2930: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:03.258516: Step 2940: Train accuracy = 90.0%
2017-09-09 16:27:03.258557: Step 2940: Cross entropy = 0.336034
2017-09-09 16:27:03.269954: Step 2940: Validation accuracy = 60.0% (N=30)
2017-09-09 16:27:03.388262: Step 2950: Train accuracy = 83.3%
2017-09-09 16:27:03.388305: Step 2950: Cross entropy = 0.448142
2017-09-09 16:27:03.399882: Step 2950: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:03.518252: Step 2960: Train accuracy = 70.0%
2017-09-09 16:27:03.518294: Step 2960: Cross entropy = 0.619627
2017-09-09 16:27:03.529782: Step 2960: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:03.649231: Step 2970: Train accuracy = 90.0%
2017-09-09 16:27:03.649273: Step 2970: Cross entropy = 0.339788
2017-09-09 16:27:03.660975: Step 2970: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:03.781119: Step 2980: Train accuracy = 70.0%
2017-09-09 16:27:03.781162: Step 2980: Cross entropy = 0.443428
2017-09-09 16:27:03.792861: Step 2980: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:03.914402: Step 2990: Train accuracy = 73.3%
2017-09-09 16:27:03.914443: Step 2990: Cross entropy = 0.528709
2017-09-09 16:27:03.926098: Step 2990: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:04.048058: Step 3000: Train accuracy = 80.0%
2017-09-09 16:27:04.048100: Step 3000: Cross entropy = 0.384838
2017-09-09 16:27:04.059787: Step 3000: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:04.181648: Step 3010: Train accuracy = 83.3%
2017-09-09 16:27:04.181690: Step 3010: Cross entropy = 0.374421
2017-09-09 16:27:04.193343: Step 3010: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:04.315077: Step 3020: Train accuracy = 83.3%
2017-09-09 16:27:04.315119: Step 3020: Cross entropy = 0.411565
2017-09-09 16:27:04.326876: Step 3020: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:04.447827: Step 3030: Train accuracy = 83.3%
2017-09-09 16:27:04.447869: Step 3030: Cross entropy = 0.409271
2017-09-09 16:27:04.459487: Step 3030: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:04.580333: Step 3040: Train accuracy = 80.0%
2017-09-09 16:27:04.580376: Step 3040: Cross entropy = 0.502441
2017-09-09 16:27:04.592042: Step 3040: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:04.712643: Step 3050: Train accuracy = 66.7%
2017-09-09 16:27:04.712685: Step 3050: Cross entropy = 0.543033
2017-09-09 16:27:04.724367: Step 3050: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:04.845235: Step 3060: Train accuracy = 83.3%
2017-09-09 16:27:04.845279: Step 3060: Cross entropy = 0.399127
2017-09-09 16:27:04.857250: Step 3060: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:04.977935: Step 3070: Train accuracy = 70.0%
2017-09-09 16:27:04.977978: Step 3070: Cross entropy = 0.510624
2017-09-09 16:27:04.989638: Step 3070: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:05.108287: Step 3080: Train accuracy = 70.0%
2017-09-09 16:27:05.108328: Step 3080: Cross entropy = 0.559952
2017-09-09 16:27:05.119826: Step 3080: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:05.238283: Step 3090: Train accuracy = 90.0%
2017-09-09 16:27:05.238325: Step 3090: Cross entropy = 0.366717
2017-09-09 16:27:05.249799: Step 3090: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:05.370449: Step 3100: Train accuracy = 86.7%
2017-09-09 16:27:05.370490: Step 3100: Cross entropy = 0.338221
2017-09-09 16:27:05.382113: Step 3100: Validation accuracy = 53.3% (N=30)
2017-09-09 16:27:05.500270: Step 3110: Train accuracy = 80.0%
2017-09-09 16:27:05.500310: Step 3110: Cross entropy = 0.439623
2017-09-09 16:27:05.511883: Step 3110: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:05.630852: Step 3120: Train accuracy = 80.0%
2017-09-09 16:27:05.630894: Step 3120: Cross entropy = 0.446581
2017-09-09 16:27:05.642406: Step 3120: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:05.760720: Step 3130: Train accuracy = 86.7%
2017-09-09 16:27:05.760760: Step 3130: Cross entropy = 0.403863
2017-09-09 16:27:05.772165: Step 3130: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:05.890043: Step 3140: Train accuracy = 63.3%
2017-09-09 16:27:05.890086: Step 3140: Cross entropy = 0.551226
2017-09-09 16:27:05.901618: Step 3140: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:06.019911: Step 3150: Train accuracy = 90.0%
2017-09-09 16:27:06.019952: Step 3150: Cross entropy = 0.361141
2017-09-09 16:27:06.031366: Step 3150: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:06.150469: Step 3160: Train accuracy = 90.0%
2017-09-09 16:27:06.150510: Step 3160: Cross entropy = 0.333747
2017-09-09 16:27:06.161929: Step 3160: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:06.280756: Step 3170: Train accuracy = 80.0%
2017-09-09 16:27:06.280798: Step 3170: Cross entropy = 0.362351
2017-09-09 16:27:06.292184: Step 3170: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:06.412506: Step 3180: Train accuracy = 90.0%
2017-09-09 16:27:06.412546: Step 3180: Cross entropy = 0.397013
2017-09-09 16:27:06.424098: Step 3180: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:06.544127: Step 3190: Train accuracy = 73.3%
2017-09-09 16:27:06.544170: Step 3190: Cross entropy = 0.469318
2017-09-09 16:27:06.555997: Step 3190: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:06.676991: Step 3200: Train accuracy = 90.0%
2017-09-09 16:27:06.677033: Step 3200: Cross entropy = 0.335517
2017-09-09 16:27:06.688920: Step 3200: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:06.811669: Step 3210: Train accuracy = 83.3%
2017-09-09 16:27:06.811723: Step 3210: Cross entropy = 0.429158
2017-09-09 16:27:06.823301: Step 3210: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:06.944766: Step 3220: Train accuracy = 83.3%
2017-09-09 16:27:06.944810: Step 3220: Cross entropy = 0.444888
2017-09-09 16:27:06.956410: Step 3220: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:07.077234: Step 3230: Train accuracy = 86.7%
2017-09-09 16:27:07.077279: Step 3230: Cross entropy = 0.451521
2017-09-09 16:27:07.088833: Step 3230: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:07.211186: Step 3240: Train accuracy = 90.0%
2017-09-09 16:27:07.211230: Step 3240: Cross entropy = 0.300159
2017-09-09 16:27:07.222882: Step 3240: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:07.344428: Step 3250: Train accuracy = 90.0%
2017-09-09 16:27:07.344470: Step 3250: Cross entropy = 0.262612
2017-09-09 16:27:07.356210: Step 3250: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:07.477122: Step 3260: Train accuracy = 96.7%
2017-09-09 16:27:07.477167: Step 3260: Cross entropy = 0.279544
2017-09-09 16:27:07.488853: Step 3260: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:07.609071: Step 3270: Train accuracy = 83.3%
2017-09-09 16:27:07.609113: Step 3270: Cross entropy = 0.412244
2017-09-09 16:27:07.620623: Step 3270: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:07.741060: Step 3280: Train accuracy = 83.3%
2017-09-09 16:27:07.741103: Step 3280: Cross entropy = 0.407389
2017-09-09 16:27:07.752611: Step 3280: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:07.871463: Step 3290: Train accuracy = 76.7%
2017-09-09 16:27:07.871503: Step 3290: Cross entropy = 0.432720
2017-09-09 16:27:07.882974: Step 3290: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:08.006681: Step 3300: Train accuracy = 90.0%
2017-09-09 16:27:08.006724: Step 3300: Cross entropy = 0.358249
2017-09-09 16:27:08.018328: Step 3300: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:08.136977: Step 3310: Train accuracy = 80.0%
2017-09-09 16:27:08.137018: Step 3310: Cross entropy = 0.365123
2017-09-09 16:27:08.148451: Step 3310: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:08.266474: Step 3320: Train accuracy = 73.3%
2017-09-09 16:27:08.266515: Step 3320: Cross entropy = 0.500827
2017-09-09 16:27:08.277947: Step 3320: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:08.395499: Step 3330: Train accuracy = 83.3%
2017-09-09 16:27:08.395539: Step 3330: Cross entropy = 0.355614
2017-09-09 16:27:08.406942: Step 3330: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:08.525696: Step 3340: Train accuracy = 86.7%
2017-09-09 16:27:08.525737: Step 3340: Cross entropy = 0.411866
2017-09-09 16:27:08.537276: Step 3340: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:08.656569: Step 3350: Train accuracy = 86.7%
2017-09-09 16:27:08.656612: Step 3350: Cross entropy = 0.321446
2017-09-09 16:27:08.668019: Step 3350: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:08.787325: Step 3360: Train accuracy = 80.0%
2017-09-09 16:27:08.787368: Step 3360: Cross entropy = 0.391160
2017-09-09 16:27:08.798927: Step 3360: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:08.917637: Step 3370: Train accuracy = 90.0%
2017-09-09 16:27:08.917680: Step 3370: Cross entropy = 0.372096
2017-09-09 16:27:08.929198: Step 3370: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:09.048849: Step 3380: Train accuracy = 76.7%
2017-09-09 16:27:09.048891: Step 3380: Cross entropy = 0.372040
2017-09-09 16:27:09.060447: Step 3380: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:09.181788: Step 3390: Train accuracy = 80.0%
2017-09-09 16:27:09.181830: Step 3390: Cross entropy = 0.519890
2017-09-09 16:27:09.193432: Step 3390: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:09.312445: Step 3400: Train accuracy = 76.7%
2017-09-09 16:27:09.312486: Step 3400: Cross entropy = 0.447765
2017-09-09 16:27:09.323841: Step 3400: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:09.442573: Step 3410: Train accuracy = 83.3%
2017-09-09 16:27:09.442615: Step 3410: Cross entropy = 0.492055
2017-09-09 16:27:09.454014: Step 3410: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:09.572438: Step 3420: Train accuracy = 80.0%
2017-09-09 16:27:09.572480: Step 3420: Cross entropy = 0.448725
2017-09-09 16:27:09.583983: Step 3420: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:09.702403: Step 3430: Train accuracy = 76.7%
2017-09-09 16:27:09.702446: Step 3430: Cross entropy = 0.466883
2017-09-09 16:27:09.713878: Step 3430: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:09.832546: Step 3440: Train accuracy = 83.3%
2017-09-09 16:27:09.832591: Step 3440: Cross entropy = 0.434905
2017-09-09 16:27:09.844029: Step 3440: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:09.962174: Step 3450: Train accuracy = 80.0%
2017-09-09 16:27:09.962216: Step 3450: Cross entropy = 0.488501
2017-09-09 16:27:09.973745: Step 3450: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:10.092862: Step 3460: Train accuracy = 93.3%
2017-09-09 16:27:10.092903: Step 3460: Cross entropy = 0.295833
2017-09-09 16:27:10.104281: Step 3460: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:10.223264: Step 3470: Train accuracy = 80.0%
2017-09-09 16:27:10.223307: Step 3470: Cross entropy = 0.413247
2017-09-09 16:27:10.234729: Step 3470: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:10.353038: Step 3480: Train accuracy = 90.0%
2017-09-09 16:27:10.353079: Step 3480: Cross entropy = 0.369313
2017-09-09 16:27:10.364701: Step 3480: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:10.484746: Step 3490: Train accuracy = 83.3%
2017-09-09 16:27:10.484787: Step 3490: Cross entropy = 0.338596
2017-09-09 16:27:10.496193: Step 3490: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:10.614795: Step 3500: Train accuracy = 83.3%
2017-09-09 16:27:10.614837: Step 3500: Cross entropy = 0.385079
2017-09-09 16:27:10.626359: Step 3500: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:10.745020: Step 3510: Train accuracy = 83.3%
2017-09-09 16:27:10.745062: Step 3510: Cross entropy = 0.437702
2017-09-09 16:27:10.756760: Step 3510: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:10.875619: Step 3520: Train accuracy = 76.7%
2017-09-09 16:27:10.875662: Step 3520: Cross entropy = 0.508390
2017-09-09 16:27:10.887512: Step 3520: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:11.009924: Step 3530: Train accuracy = 83.3%
2017-09-09 16:27:11.009966: Step 3530: Cross entropy = 0.375668
2017-09-09 16:27:11.021737: Step 3530: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:11.143642: Step 3540: Train accuracy = 76.7%
2017-09-09 16:27:11.143694: Step 3540: Cross entropy = 0.468009
2017-09-09 16:27:11.155732: Step 3540: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:11.277213: Step 3550: Train accuracy = 86.7%
2017-09-09 16:27:11.277256: Step 3550: Cross entropy = 0.355543
2017-09-09 16:27:11.288967: Step 3550: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:11.410957: Step 3560: Train accuracy = 86.7%
2017-09-09 16:27:11.411001: Step 3560: Cross entropy = 0.405446
2017-09-09 16:27:11.422835: Step 3560: Validation accuracy = 60.0% (N=30)
2017-09-09 16:27:11.544908: Step 3570: Train accuracy = 80.0%
2017-09-09 16:27:11.544952: Step 3570: Cross entropy = 0.504976
2017-09-09 16:27:11.556888: Step 3570: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:11.677565: Step 3580: Train accuracy = 76.7%
2017-09-09 16:27:11.677610: Step 3580: Cross entropy = 0.460904
2017-09-09 16:27:11.689239: Step 3580: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:11.810648: Step 3590: Train accuracy = 90.0%
2017-09-09 16:27:11.810691: Step 3590: Cross entropy = 0.336075
2017-09-09 16:27:11.823448: Step 3590: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:11.947014: Step 3600: Train accuracy = 86.7%
2017-09-09 16:27:11.947055: Step 3600: Cross entropy = 0.418314
2017-09-09 16:27:11.959453: Step 3600: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:12.083819: Step 3610: Train accuracy = 80.0%
2017-09-09 16:27:12.083863: Step 3610: Cross entropy = 0.394825
2017-09-09 16:27:12.096469: Step 3610: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:12.218769: Step 3620: Train accuracy = 83.3%
2017-09-09 16:27:12.218811: Step 3620: Cross entropy = 0.336998
2017-09-09 16:27:12.230587: Step 3620: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:12.351334: Step 3630: Train accuracy = 83.3%
2017-09-09 16:27:12.351376: Step 3630: Cross entropy = 0.481563
2017-09-09 16:27:12.362850: Step 3630: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:12.484509: Step 3640: Train accuracy = 73.3%
2017-09-09 16:27:12.484550: Step 3640: Cross entropy = 0.497507
2017-09-09 16:27:12.496139: Step 3640: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:12.615313: Step 3650: Train accuracy = 86.7%
2017-09-09 16:27:12.615355: Step 3650: Cross entropy = 0.354148
2017-09-09 16:27:12.626852: Step 3650: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:12.745826: Step 3660: Train accuracy = 86.7%
2017-09-09 16:27:12.745869: Step 3660: Cross entropy = 0.353031
2017-09-09 16:27:12.757378: Step 3660: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:12.876019: Step 3670: Train accuracy = 80.0%
2017-09-09 16:27:12.876061: Step 3670: Cross entropy = 0.418472
2017-09-09 16:27:12.887597: Step 3670: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:13.009275: Step 3680: Train accuracy = 86.7%
2017-09-09 16:27:13.009317: Step 3680: Cross entropy = 0.300359
2017-09-09 16:27:13.020686: Step 3680: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:13.139844: Step 3690: Train accuracy = 80.0%
2017-09-09 16:27:13.139885: Step 3690: Cross entropy = 0.425399
2017-09-09 16:27:13.151454: Step 3690: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:13.270961: Step 3700: Train accuracy = 96.7%
2017-09-09 16:27:13.271003: Step 3700: Cross entropy = 0.225872
2017-09-09 16:27:13.282476: Step 3700: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:13.403200: Step 3710: Train accuracy = 83.3%
2017-09-09 16:27:13.403242: Step 3710: Cross entropy = 0.374462
2017-09-09 16:27:13.414802: Step 3710: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:13.536443: Step 3720: Train accuracy = 73.3%
2017-09-09 16:27:13.536487: Step 3720: Cross entropy = 0.499513
2017-09-09 16:27:13.548107: Step 3720: Validation accuracy = 56.7% (N=30)
2017-09-09 16:27:13.670426: Step 3730: Train accuracy = 83.3%
2017-09-09 16:27:13.670468: Step 3730: Cross entropy = 0.430237
2017-09-09 16:27:13.682214: Step 3730: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:13.804383: Step 3740: Train accuracy = 86.7%
2017-09-09 16:27:13.804426: Step 3740: Cross entropy = 0.362603
2017-09-09 16:27:13.816376: Step 3740: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:13.938923: Step 3750: Train accuracy = 73.3%
2017-09-09 16:27:13.938965: Step 3750: Cross entropy = 0.468333
2017-09-09 16:27:13.950822: Step 3750: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:14.074130: Step 3760: Train accuracy = 90.0%
2017-09-09 16:27:14.074174: Step 3760: Cross entropy = 0.355234
2017-09-09 16:27:14.085969: Step 3760: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:14.207881: Step 3770: Train accuracy = 80.0%
2017-09-09 16:27:14.207923: Step 3770: Cross entropy = 0.432870
2017-09-09 16:27:14.219637: Step 3770: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:14.340981: Step 3780: Train accuracy = 90.0%
2017-09-09 16:27:14.341021: Step 3780: Cross entropy = 0.315946
2017-09-09 16:27:14.352657: Step 3780: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:14.473330: Step 3790: Train accuracy = 73.3%
2017-09-09 16:27:14.473372: Step 3790: Cross entropy = 0.454053
2017-09-09 16:27:14.484965: Step 3790: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:14.617301: Step 3800: Train accuracy = 76.7%
2017-09-09 16:27:14.617344: Step 3800: Cross entropy = 0.514598
2017-09-09 16:27:14.629017: Step 3800: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:14.761500: Step 3810: Train accuracy = 83.3%
2017-09-09 16:27:14.761542: Step 3810: Cross entropy = 0.348290
2017-09-09 16:27:14.773213: Step 3810: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:14.910757: Step 3820: Train accuracy = 76.7%
2017-09-09 16:27:14.910799: Step 3820: Cross entropy = 0.500413
2017-09-09 16:27:14.922483: Step 3820: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:15.043553: Step 3830: Train accuracy = 80.0%
2017-09-09 16:27:15.043595: Step 3830: Cross entropy = 0.405546
2017-09-09 16:27:15.055365: Step 3830: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:15.174989: Step 3840: Train accuracy = 83.3%
2017-09-09 16:27:15.175031: Step 3840: Cross entropy = 0.361296
2017-09-09 16:27:15.186614: Step 3840: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:15.306139: Step 3850: Train accuracy = 66.7%
2017-09-09 16:27:15.306180: Step 3850: Cross entropy = 0.516491
2017-09-09 16:27:15.317744: Step 3850: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:15.437477: Step 3860: Train accuracy = 83.3%
2017-09-09 16:27:15.437520: Step 3860: Cross entropy = 0.401852
2017-09-09 16:27:15.449315: Step 3860: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:15.569404: Step 3870: Train accuracy = 90.0%
2017-09-09 16:27:15.569446: Step 3870: Cross entropy = 0.306215
2017-09-09 16:27:15.581028: Step 3870: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:15.701952: Step 3880: Train accuracy = 80.0%
2017-09-09 16:27:15.701995: Step 3880: Cross entropy = 0.418933
2017-09-09 16:27:15.713674: Step 3880: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:15.833940: Step 3890: Train accuracy = 73.3%
2017-09-09 16:27:15.833982: Step 3890: Cross entropy = 0.425488
2017-09-09 16:27:15.845657: Step 3890: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:15.965146: Step 3900: Train accuracy = 86.7%
2017-09-09 16:27:15.965188: Step 3900: Cross entropy = 0.359233
2017-09-09 16:27:15.976812: Step 3900: Validation accuracy = 56.7% (N=30)
2017-09-09 16:27:16.096118: Step 3910: Train accuracy = 76.7%
2017-09-09 16:27:16.096163: Step 3910: Cross entropy = 0.463819
2017-09-09 16:27:16.107620: Step 3910: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:16.227674: Step 3920: Train accuracy = 83.3%
2017-09-09 16:27:16.227726: Step 3920: Cross entropy = 0.482213
2017-09-09 16:27:16.240069: Step 3920: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:16.364892: Step 3930: Train accuracy = 76.7%
2017-09-09 16:27:16.364934: Step 3930: Cross entropy = 0.416186
2017-09-09 16:27:16.376766: Step 3930: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:16.496062: Step 3940: Train accuracy = 93.3%
2017-09-09 16:27:16.496104: Step 3940: Cross entropy = 0.240720
2017-09-09 16:27:16.507880: Step 3940: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:16.627952: Step 3950: Train accuracy = 90.0%
2017-09-09 16:27:16.627994: Step 3950: Cross entropy = 0.288480
2017-09-09 16:27:16.639636: Step 3950: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:16.759672: Step 3960: Train accuracy = 80.0%
2017-09-09 16:27:16.759721: Step 3960: Cross entropy = 0.503809
2017-09-09 16:27:16.771478: Step 3960: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:16.891416: Step 3970: Train accuracy = 83.3%
2017-09-09 16:27:16.891457: Step 3970: Cross entropy = 0.441639
2017-09-09 16:27:16.902987: Step 3970: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:17.021540: Step 3980: Train accuracy = 83.3%
2017-09-09 16:27:17.021581: Step 3980: Cross entropy = 0.344099
2017-09-09 16:27:17.033063: Step 3980: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:17.152000: Step 3990: Train accuracy = 83.3%
2017-09-09 16:27:17.152042: Step 3990: Cross entropy = 0.408243
2017-09-09 16:27:17.163478: Step 3990: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:17.282029: Step 4000: Train accuracy = 86.7%
2017-09-09 16:27:17.282070: Step 4000: Cross entropy = 0.380924
2017-09-09 16:27:17.293710: Step 4000: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:17.412188: Step 4010: Train accuracy = 90.0%
2017-09-09 16:27:17.412231: Step 4010: Cross entropy = 0.331914
2017-09-09 16:27:17.423637: Step 4010: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:17.542862: Step 4020: Train accuracy = 86.7%
2017-09-09 16:27:17.542905: Step 4020: Cross entropy = 0.395970
2017-09-09 16:27:17.554703: Step 4020: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:17.676534: Step 4030: Train accuracy = 83.3%
2017-09-09 16:27:17.676575: Step 4030: Cross entropy = 0.466609
2017-09-09 16:27:17.688038: Step 4030: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:17.807086: Step 4040: Train accuracy = 86.7%
2017-09-09 16:27:17.807127: Step 4040: Cross entropy = 0.365327
2017-09-09 16:27:17.818561: Step 4040: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:17.938321: Step 4050: Train accuracy = 80.0%
2017-09-09 16:27:17.938361: Step 4050: Cross entropy = 0.401326
2017-09-09 16:27:17.950431: Step 4050: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:18.070749: Step 4060: Train accuracy = 83.3%
2017-09-09 16:27:18.070793: Step 4060: Cross entropy = 0.324864
2017-09-09 16:27:18.082356: Step 4060: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:18.200842: Step 4070: Train accuracy = 86.7%
2017-09-09 16:27:18.200884: Step 4070: Cross entropy = 0.354894
2017-09-09 16:27:18.212618: Step 4070: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:18.331247: Step 4080: Train accuracy = 80.0%
2017-09-09 16:27:18.331289: Step 4080: Cross entropy = 0.378506
2017-09-09 16:27:18.342858: Step 4080: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:18.461916: Step 4090: Train accuracy = 86.7%
2017-09-09 16:27:18.461960: Step 4090: Cross entropy = 0.409900
2017-09-09 16:27:18.473524: Step 4090: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:18.593279: Step 4100: Train accuracy = 80.0%
2017-09-09 16:27:18.593321: Step 4100: Cross entropy = 0.432132
2017-09-09 16:27:18.605053: Step 4100: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:18.725921: Step 4110: Train accuracy = 76.7%
2017-09-09 16:27:18.725963: Step 4110: Cross entropy = 0.505285
2017-09-09 16:27:18.737456: Step 4110: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:18.856211: Step 4120: Train accuracy = 76.7%
2017-09-09 16:27:18.856252: Step 4120: Cross entropy = 0.430593
2017-09-09 16:27:18.867720: Step 4120: Validation accuracy = 96.7% (N=30)
2017-09-09 16:27:18.986774: Step 4130: Train accuracy = 90.0%
2017-09-09 16:27:18.986816: Step 4130: Cross entropy = 0.270097
2017-09-09 16:27:18.998319: Step 4130: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:19.117299: Step 4140: Train accuracy = 83.3%
2017-09-09 16:27:19.117340: Step 4140: Cross entropy = 0.378198
2017-09-09 16:27:19.128933: Step 4140: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:19.247247: Step 4150: Train accuracy = 70.0%
2017-09-09 16:27:19.247289: Step 4150: Cross entropy = 0.472772
2017-09-09 16:27:19.258893: Step 4150: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:19.377723: Step 4160: Train accuracy = 80.0%
2017-09-09 16:27:19.377766: Step 4160: Cross entropy = 0.468975
2017-09-09 16:27:19.389650: Step 4160: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:19.509197: Step 4170: Train accuracy = 76.7%
2017-09-09 16:27:19.509238: Step 4170: Cross entropy = 0.471869
2017-09-09 16:27:19.521582: Step 4170: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:19.639875: Step 4180: Train accuracy = 83.3%
2017-09-09 16:27:19.639916: Step 4180: Cross entropy = 0.403135
2017-09-09 16:27:19.651702: Step 4180: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:19.770996: Step 4190: Train accuracy = 86.7%
2017-09-09 16:27:19.771039: Step 4190: Cross entropy = 0.311497
2017-09-09 16:27:19.782523: Step 4190: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:19.901453: Step 4200: Train accuracy = 83.3%
2017-09-09 16:27:19.901496: Step 4200: Cross entropy = 0.410583
2017-09-09 16:27:19.913175: Step 4200: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:20.032683: Step 4210: Train accuracy = 86.7%
2017-09-09 16:27:20.032725: Step 4210: Cross entropy = 0.409220
2017-09-09 16:27:20.044172: Step 4210: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:20.162592: Step 4220: Train accuracy = 80.0%
2017-09-09 16:27:20.162634: Step 4220: Cross entropy = 0.443062
2017-09-09 16:27:20.174352: Step 4220: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:20.292237: Step 4230: Train accuracy = 70.0%
2017-09-09 16:27:20.292279: Step 4230: Cross entropy = 0.557693
2017-09-09 16:27:20.303758: Step 4230: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:20.422091: Step 4240: Train accuracy = 83.3%
2017-09-09 16:27:20.422133: Step 4240: Cross entropy = 0.380919
2017-09-09 16:27:20.433998: Step 4240: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:20.552206: Step 4250: Train accuracy = 70.0%
2017-09-09 16:27:20.552249: Step 4250: Cross entropy = 0.553271
2017-09-09 16:27:20.563669: Step 4250: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:20.682957: Step 4260: Train accuracy = 80.0%
2017-09-09 16:27:20.682998: Step 4260: Cross entropy = 0.408733
2017-09-09 16:27:20.694545: Step 4260: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:20.812759: Step 4270: Train accuracy = 83.3%
2017-09-09 16:27:20.812800: Step 4270: Cross entropy = 0.369762
2017-09-09 16:27:20.824335: Step 4270: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:20.943481: Step 4280: Train accuracy = 93.3%
2017-09-09 16:27:20.943522: Step 4280: Cross entropy = 0.293897
2017-09-09 16:27:20.955063: Step 4280: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:21.073639: Step 4290: Train accuracy = 83.3%
2017-09-09 16:27:21.073681: Step 4290: Cross entropy = 0.441228
2017-09-09 16:27:21.085222: Step 4290: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:21.204145: Step 4300: Train accuracy = 90.0%
2017-09-09 16:27:21.204187: Step 4300: Cross entropy = 0.294294
2017-09-09 16:27:21.215856: Step 4300: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:21.334070: Step 4310: Train accuracy = 80.0%
2017-09-09 16:27:21.334112: Step 4310: Cross entropy = 0.343554
2017-09-09 16:27:21.345716: Step 4310: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:21.464742: Step 4320: Train accuracy = 90.0%
2017-09-09 16:27:21.464785: Step 4320: Cross entropy = 0.365608
2017-09-09 16:27:21.476256: Step 4320: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:21.596121: Step 4330: Train accuracy = 76.7%
2017-09-09 16:27:21.596162: Step 4330: Cross entropy = 0.438234
2017-09-09 16:27:21.607736: Step 4330: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:21.726505: Step 4340: Train accuracy = 83.3%
2017-09-09 16:27:21.726548: Step 4340: Cross entropy = 0.421842
2017-09-09 16:27:21.738107: Step 4340: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:21.856000: Step 4350: Train accuracy = 83.3%
2017-09-09 16:27:21.856042: Step 4350: Cross entropy = 0.384884
2017-09-09 16:27:21.867553: Step 4350: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:21.986666: Step 4360: Train accuracy = 83.3%
2017-09-09 16:27:21.986709: Step 4360: Cross entropy = 0.359532
2017-09-09 16:27:21.998293: Step 4360: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:22.116750: Step 4370: Train accuracy = 73.3%
2017-09-09 16:27:22.116792: Step 4370: Cross entropy = 0.469228
2017-09-09 16:27:22.128285: Step 4370: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:22.246949: Step 4380: Train accuracy = 80.0%
2017-09-09 16:27:22.246994: Step 4380: Cross entropy = 0.447475
2017-09-09 16:27:22.258571: Step 4380: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:22.377093: Step 4390: Train accuracy = 93.3%
2017-09-09 16:27:22.377135: Step 4390: Cross entropy = 0.287605
2017-09-09 16:27:22.388722: Step 4390: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:22.508336: Step 4400: Train accuracy = 86.7%
2017-09-09 16:27:22.508378: Step 4400: Cross entropy = 0.346650
2017-09-09 16:27:22.519969: Step 4400: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:22.637854: Step 4410: Train accuracy = 90.0%
2017-09-09 16:27:22.637897: Step 4410: Cross entropy = 0.331366
2017-09-09 16:27:22.649326: Step 4410: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:22.768662: Step 4420: Train accuracy = 83.3%
2017-09-09 16:27:22.768704: Step 4420: Cross entropy = 0.314580
2017-09-09 16:27:22.780426: Step 4420: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:22.899172: Step 4430: Train accuracy = 80.0%
2017-09-09 16:27:22.899212: Step 4430: Cross entropy = 0.405440
2017-09-09 16:27:22.910671: Step 4430: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:23.030911: Step 4440: Train accuracy = 90.0%
2017-09-09 16:27:23.030953: Step 4440: Cross entropy = 0.336589
2017-09-09 16:27:23.042548: Step 4440: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:23.161404: Step 4450: Train accuracy = 76.7%
2017-09-09 16:27:23.161447: Step 4450: Cross entropy = 0.438848
2017-09-09 16:27:23.173672: Step 4450: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:23.292749: Step 4460: Train accuracy = 86.7%
2017-09-09 16:27:23.292790: Step 4460: Cross entropy = 0.390445
2017-09-09 16:27:23.304250: Step 4460: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:23.424272: Step 4470: Train accuracy = 90.0%
2017-09-09 16:27:23.424314: Step 4470: Cross entropy = 0.360639
2017-09-09 16:27:23.435831: Step 4470: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:23.553948: Step 4480: Train accuracy = 86.7%
2017-09-09 16:27:23.553991: Step 4480: Cross entropy = 0.448066
2017-09-09 16:27:23.565484: Step 4480: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:23.685238: Step 4490: Train accuracy = 90.0%
2017-09-09 16:27:23.685281: Step 4490: Cross entropy = 0.430672
2017-09-09 16:27:23.696795: Step 4490: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:23.814874: Step 4500: Train accuracy = 90.0%
2017-09-09 16:27:23.814916: Step 4500: Cross entropy = 0.284885
2017-09-09 16:27:23.826530: Step 4500: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:23.944474: Step 4510: Train accuracy = 83.3%
2017-09-09 16:27:23.944514: Step 4510: Cross entropy = 0.342802
2017-09-09 16:27:23.955929: Step 4510: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:24.075305: Step 4520: Train accuracy = 96.7%
2017-09-09 16:27:24.075348: Step 4520: Cross entropy = 0.254368
2017-09-09 16:27:24.086734: Step 4520: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:24.204513: Step 4530: Train accuracy = 83.3%
2017-09-09 16:27:24.204555: Step 4530: Cross entropy = 0.418823
2017-09-09 16:27:24.216052: Step 4530: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:24.334372: Step 4540: Train accuracy = 76.7%
2017-09-09 16:27:24.334414: Step 4540: Cross entropy = 0.440015
2017-09-09 16:27:24.345996: Step 4540: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:24.464095: Step 4550: Train accuracy = 76.7%
2017-09-09 16:27:24.464138: Step 4550: Cross entropy = 0.487594
2017-09-09 16:27:24.475592: Step 4550: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:24.595853: Step 4560: Train accuracy = 83.3%
2017-09-09 16:27:24.595895: Step 4560: Cross entropy = 0.393820
2017-09-09 16:27:24.607481: Step 4560: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:24.726320: Step 4570: Train accuracy = 76.7%
2017-09-09 16:27:24.726363: Step 4570: Cross entropy = 0.466759
2017-09-09 16:27:24.737828: Step 4570: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:24.857704: Step 4580: Train accuracy = 86.7%
2017-09-09 16:27:24.857746: Step 4580: Cross entropy = 0.355925
2017-09-09 16:27:24.869934: Step 4580: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:24.989636: Step 4590: Train accuracy = 83.3%
2017-09-09 16:27:24.989678: Step 4590: Cross entropy = 0.449648
2017-09-09 16:27:25.001437: Step 4590: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:25.120292: Step 4600: Train accuracy = 80.0%
2017-09-09 16:27:25.120333: Step 4600: Cross entropy = 0.368427
2017-09-09 16:27:25.131801: Step 4600: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:25.250468: Step 4610: Train accuracy = 83.3%
2017-09-09 16:27:25.250510: Step 4610: Cross entropy = 0.432831
2017-09-09 16:27:25.261992: Step 4610: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:25.380565: Step 4620: Train accuracy = 73.3%
2017-09-09 16:27:25.380607: Step 4620: Cross entropy = 0.474654
2017-09-09 16:27:25.392041: Step 4620: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:25.510270: Step 4630: Train accuracy = 83.3%
2017-09-09 16:27:25.510312: Step 4630: Cross entropy = 0.416909
2017-09-09 16:27:25.521892: Step 4630: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:25.640690: Step 4640: Train accuracy = 83.3%
2017-09-09 16:27:25.640732: Step 4640: Cross entropy = 0.404255
2017-09-09 16:27:25.652345: Step 4640: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:25.770959: Step 4650: Train accuracy = 86.7%
2017-09-09 16:27:25.771002: Step 4650: Cross entropy = 0.401003
2017-09-09 16:27:25.782514: Step 4650: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:25.902473: Step 4660: Train accuracy = 90.0%
2017-09-09 16:27:25.902516: Step 4660: Cross entropy = 0.326168
2017-09-09 16:27:25.914010: Step 4660: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:26.031979: Step 4670: Train accuracy = 93.3%
2017-09-09 16:27:26.032021: Step 4670: Cross entropy = 0.309637
2017-09-09 16:27:26.043586: Step 4670: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:26.162255: Step 4680: Train accuracy = 93.3%
2017-09-09 16:27:26.162297: Step 4680: Cross entropy = 0.278357
2017-09-09 16:27:26.173919: Step 4680: Validation accuracy = 56.7% (N=30)
2017-09-09 16:27:26.291925: Step 4690: Train accuracy = 83.3%
2017-09-09 16:27:26.291967: Step 4690: Cross entropy = 0.384067
2017-09-09 16:27:26.303388: Step 4690: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:26.424942: Step 4700: Train accuracy = 76.7%
2017-09-09 16:27:26.424985: Step 4700: Cross entropy = 0.429303
2017-09-09 16:27:26.437526: Step 4700: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:26.555466: Step 4710: Train accuracy = 66.7%
2017-09-09 16:27:26.555508: Step 4710: Cross entropy = 0.581484
2017-09-09 16:27:26.567122: Step 4710: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:26.685334: Step 4720: Train accuracy = 90.0%
2017-09-09 16:27:26.685377: Step 4720: Cross entropy = 0.307986
2017-09-09 16:27:26.696922: Step 4720: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:26.814943: Step 4730: Train accuracy = 86.7%
2017-09-09 16:27:26.814985: Step 4730: Cross entropy = 0.343864
2017-09-09 16:27:26.827177: Step 4730: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:26.946746: Step 4740: Train accuracy = 90.0%
2017-09-09 16:27:26.946787: Step 4740: Cross entropy = 0.418093
2017-09-09 16:27:26.958232: Step 4740: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:27.076601: Step 4750: Train accuracy = 90.0%
2017-09-09 16:27:27.076642: Step 4750: Cross entropy = 0.362778
2017-09-09 16:27:27.088142: Step 4750: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:27.207621: Step 4760: Train accuracy = 86.7%
2017-09-09 16:27:27.207664: Step 4760: Cross entropy = 0.407437
2017-09-09 16:27:27.219150: Step 4760: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:27.337941: Step 4770: Train accuracy = 63.3%
2017-09-09 16:27:27.337983: Step 4770: Cross entropy = 0.586219
2017-09-09 16:27:27.349535: Step 4770: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:27.468023: Step 4780: Train accuracy = 80.0%
2017-09-09 16:27:27.468065: Step 4780: Cross entropy = 0.433823
2017-09-09 16:27:27.479686: Step 4780: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:27.597941: Step 4790: Train accuracy = 96.7%
2017-09-09 16:27:27.597983: Step 4790: Cross entropy = 0.294976
2017-09-09 16:27:27.609393: Step 4790: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:27.728412: Step 4800: Train accuracy = 93.3%
2017-09-09 16:27:27.728454: Step 4800: Cross entropy = 0.356611
2017-09-09 16:27:27.739954: Step 4800: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:27.858378: Step 4810: Train accuracy = 93.3%
2017-09-09 16:27:27.858421: Step 4810: Cross entropy = 0.289532
2017-09-09 16:27:27.869957: Step 4810: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:27.988274: Step 4820: Train accuracy = 83.3%
2017-09-09 16:27:27.988317: Step 4820: Cross entropy = 0.302314
2017-09-09 16:27:27.999887: Step 4820: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:28.117851: Step 4830: Train accuracy = 76.7%
2017-09-09 16:27:28.117892: Step 4830: Cross entropy = 0.521016
2017-09-09 16:27:28.129413: Step 4830: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:28.248026: Step 4840: Train accuracy = 83.3%
2017-09-09 16:27:28.248068: Step 4840: Cross entropy = 0.362696
2017-09-09 16:27:28.260381: Step 4840: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:28.379603: Step 4850: Train accuracy = 86.7%
2017-09-09 16:27:28.379644: Step 4850: Cross entropy = 0.298462
2017-09-09 16:27:28.391190: Step 4850: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:28.509647: Step 4860: Train accuracy = 96.7%
2017-09-09 16:27:28.509689: Step 4860: Cross entropy = 0.277119
2017-09-09 16:27:28.521281: Step 4860: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:28.639631: Step 4870: Train accuracy = 83.3%
2017-09-09 16:27:28.639673: Step 4870: Cross entropy = 0.374530
2017-09-09 16:27:28.651331: Step 4870: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:28.769875: Step 4880: Train accuracy = 93.3%
2017-09-09 16:27:28.769916: Step 4880: Cross entropy = 0.297132
2017-09-09 16:27:28.781491: Step 4880: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:28.900369: Step 4890: Train accuracy = 86.7%
2017-09-09 16:27:28.900409: Step 4890: Cross entropy = 0.324455
2017-09-09 16:27:28.911899: Step 4890: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:29.030279: Step 4900: Train accuracy = 73.3%
2017-09-09 16:27:29.030321: Step 4900: Cross entropy = 0.465772
2017-09-09 16:27:29.041850: Step 4900: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:29.159792: Step 4910: Train accuracy = 66.7%
2017-09-09 16:27:29.159834: Step 4910: Cross entropy = 0.566994
2017-09-09 16:27:29.171255: Step 4910: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:29.289609: Step 4920: Train accuracy = 83.3%
2017-09-09 16:27:29.289651: Step 4920: Cross entropy = 0.371877
2017-09-09 16:27:29.301252: Step 4920: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:29.419492: Step 4930: Train accuracy = 73.3%
2017-09-09 16:27:29.419534: Step 4930: Cross entropy = 0.454801
2017-09-09 16:27:29.432102: Step 4930: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:29.552468: Step 4940: Train accuracy = 83.3%
2017-09-09 16:27:29.552508: Step 4940: Cross entropy = 0.377748
2017-09-09 16:27:29.564111: Step 4940: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:29.683047: Step 4950: Train accuracy = 83.3%
2017-09-09 16:27:29.683089: Step 4950: Cross entropy = 0.441607
2017-09-09 16:27:29.694760: Step 4950: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:29.812575: Step 4960: Train accuracy = 83.3%
2017-09-09 16:27:29.812617: Step 4960: Cross entropy = 0.357612
2017-09-09 16:27:29.824573: Step 4960: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:29.943023: Step 4970: Train accuracy = 76.7%
2017-09-09 16:27:29.943065: Step 4970: Cross entropy = 0.504486
2017-09-09 16:27:29.954779: Step 4970: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:30.073274: Step 4980: Train accuracy = 90.0%
2017-09-09 16:27:30.073317: Step 4980: Cross entropy = 0.328988
2017-09-09 16:27:30.085161: Step 4980: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:30.204606: Step 4990: Train accuracy = 93.3%
2017-09-09 16:27:30.204647: Step 4990: Cross entropy = 0.308526
2017-09-09 16:27:30.216042: Step 4990: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:30.335180: Step 5000: Train accuracy = 80.0%
2017-09-09 16:27:30.335222: Step 5000: Cross entropy = 0.389123
2017-09-09 16:27:30.346682: Step 5000: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:30.464439: Step 5010: Train accuracy = 70.0%
2017-09-09 16:27:30.464480: Step 5010: Cross entropy = 0.537859
2017-09-09 16:27:30.475978: Step 5010: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:30.594936: Step 5020: Train accuracy = 96.7%
2017-09-09 16:27:30.594980: Step 5020: Cross entropy = 0.272707
2017-09-09 16:27:30.606573: Step 5020: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:30.727061: Step 5030: Train accuracy = 86.7%
2017-09-09 16:27:30.727102: Step 5030: Cross entropy = 0.375762
2017-09-09 16:27:30.738604: Step 5030: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:30.857271: Step 5040: Train accuracy = 76.7%
2017-09-09 16:27:30.857314: Step 5040: Cross entropy = 0.387941
2017-09-09 16:27:30.869048: Step 5040: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:30.987755: Step 5050: Train accuracy = 83.3%
2017-09-09 16:27:30.987797: Step 5050: Cross entropy = 0.415061
2017-09-09 16:27:30.999431: Step 5050: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:31.117985: Step 5060: Train accuracy = 83.3%
2017-09-09 16:27:31.118026: Step 5060: Cross entropy = 0.365815
2017-09-09 16:27:31.129486: Step 5060: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:31.248059: Step 5070: Train accuracy = 76.7%
2017-09-09 16:27:31.248102: Step 5070: Cross entropy = 0.541232
2017-09-09 16:27:31.259617: Step 5070: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:31.378179: Step 5080: Train accuracy = 86.7%
2017-09-09 16:27:31.378220: Step 5080: Cross entropy = 0.403029
2017-09-09 16:27:31.390119: Step 5080: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:31.509458: Step 5090: Train accuracy = 86.7%
2017-09-09 16:27:31.509500: Step 5090: Cross entropy = 0.322131
2017-09-09 16:27:31.521024: Step 5090: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:31.640178: Step 5100: Train accuracy = 70.0%
2017-09-09 16:27:31.640220: Step 5100: Cross entropy = 0.444487
2017-09-09 16:27:31.651741: Step 5100: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:31.769768: Step 5110: Train accuracy = 83.3%
2017-09-09 16:27:31.769811: Step 5110: Cross entropy = 0.361728
2017-09-09 16:27:31.781354: Step 5110: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:31.901409: Step 5120: Train accuracy = 90.0%
2017-09-09 16:27:31.901450: Step 5120: Cross entropy = 0.316453
2017-09-09 16:27:31.912902: Step 5120: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:32.030479: Step 5130: Train accuracy = 76.7%
2017-09-09 16:27:32.030520: Step 5130: Cross entropy = 0.447087
2017-09-09 16:27:32.042312: Step 5130: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:32.161379: Step 5140: Train accuracy = 83.3%
2017-09-09 16:27:32.161422: Step 5140: Cross entropy = 0.428867
2017-09-09 16:27:32.172996: Step 5140: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:32.290944: Step 5150: Train accuracy = 80.0%
2017-09-09 16:27:32.290985: Step 5150: Cross entropy = 0.473609
2017-09-09 16:27:32.302567: Step 5150: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:32.420885: Step 5160: Train accuracy = 76.7%
2017-09-09 16:27:32.420927: Step 5160: Cross entropy = 0.527932
2017-09-09 16:27:32.432347: Step 5160: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:32.551055: Step 5170: Train accuracy = 83.3%
2017-09-09 16:27:32.551097: Step 5170: Cross entropy = 0.400926
2017-09-09 16:27:32.562541: Step 5170: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:32.680964: Step 5180: Train accuracy = 90.0%
2017-09-09 16:27:32.681005: Step 5180: Cross entropy = 0.290761
2017-09-09 16:27:32.692469: Step 5180: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:32.812360: Step 5190: Train accuracy = 73.3%
2017-09-09 16:27:32.812403: Step 5190: Cross entropy = 0.417133
2017-09-09 16:27:32.823823: Step 5190: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:32.942743: Step 5200: Train accuracy = 73.3%
2017-09-09 16:27:32.942785: Step 5200: Cross entropy = 0.493116
2017-09-09 16:27:32.954305: Step 5200: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:33.072999: Step 5210: Train accuracy = 66.7%
2017-09-09 16:27:33.073041: Step 5210: Cross entropy = 0.590574
2017-09-09 16:27:33.084413: Step 5210: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:33.202502: Step 5220: Train accuracy = 86.7%
2017-09-09 16:27:33.202545: Step 5220: Cross entropy = 0.332105
2017-09-09 16:27:33.214124: Step 5220: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:33.333160: Step 5230: Train accuracy = 96.7%
2017-09-09 16:27:33.333201: Step 5230: Cross entropy = 0.294307
2017-09-09 16:27:33.344740: Step 5230: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:33.462871: Step 5240: Train accuracy = 86.7%
2017-09-09 16:27:33.462917: Step 5240: Cross entropy = 0.419248
2017-09-09 16:27:33.474507: Step 5240: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:33.593062: Step 5250: Train accuracy = 76.7%
2017-09-09 16:27:33.593105: Step 5250: Cross entropy = 0.411026
2017-09-09 16:27:33.604649: Step 5250: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:33.724162: Step 5260: Train accuracy = 86.7%
2017-09-09 16:27:33.724205: Step 5260: Cross entropy = 0.427530
2017-09-09 16:27:33.735775: Step 5260: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:33.853508: Step 5270: Train accuracy = 86.7%
2017-09-09 16:27:33.853549: Step 5270: Cross entropy = 0.342838
2017-09-09 16:27:33.865080: Step 5270: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:33.984015: Step 5280: Train accuracy = 80.0%
2017-09-09 16:27:33.984057: Step 5280: Cross entropy = 0.362628
2017-09-09 16:27:33.995555: Step 5280: Validation accuracy = 53.3% (N=30)
2017-09-09 16:27:34.115534: Step 5290: Train accuracy = 86.7%
2017-09-09 16:27:34.115578: Step 5290: Cross entropy = 0.424876
2017-09-09 16:27:34.127292: Step 5290: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:34.245896: Step 5300: Train accuracy = 83.3%
2017-09-09 16:27:34.245938: Step 5300: Cross entropy = 0.337686
2017-09-09 16:27:34.257509: Step 5300: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:34.376701: Step 5310: Train accuracy = 63.3%
2017-09-09 16:27:34.376743: Step 5310: Cross entropy = 0.508750
2017-09-09 16:27:34.388269: Step 5310: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:34.505998: Step 5320: Train accuracy = 96.7%
2017-09-09 16:27:34.506039: Step 5320: Cross entropy = 0.303234
2017-09-09 16:27:34.517539: Step 5320: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:34.637616: Step 5330: Train accuracy = 83.3%
2017-09-09 16:27:34.637658: Step 5330: Cross entropy = 0.451258
2017-09-09 16:27:34.649359: Step 5330: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:34.768277: Step 5340: Train accuracy = 100.0%
2017-09-09 16:27:34.768318: Step 5340: Cross entropy = 0.251815
2017-09-09 16:27:34.779862: Step 5340: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:34.898300: Step 5350: Train accuracy = 83.3%
2017-09-09 16:27:34.898345: Step 5350: Cross entropy = 0.397129
2017-09-09 16:27:34.909912: Step 5350: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:35.028231: Step 5360: Train accuracy = 76.7%
2017-09-09 16:27:35.028274: Step 5360: Cross entropy = 0.478691
2017-09-09 16:27:35.039869: Step 5360: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:35.157996: Step 5370: Train accuracy = 80.0%
2017-09-09 16:27:35.158037: Step 5370: Cross entropy = 0.450947
2017-09-09 16:27:35.169541: Step 5370: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:35.288304: Step 5380: Train accuracy = 80.0%
2017-09-09 16:27:35.288344: Step 5380: Cross entropy = 0.455930
2017-09-09 16:27:35.299894: Step 5380: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:35.418172: Step 5390: Train accuracy = 70.0%
2017-09-09 16:27:35.418215: Step 5390: Cross entropy = 0.587104
2017-09-09 16:27:35.429647: Step 5390: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:35.548423: Step 5400: Train accuracy = 93.3%
2017-09-09 16:27:35.548465: Step 5400: Cross entropy = 0.273274
2017-09-09 16:27:35.560102: Step 5400: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:35.678089: Step 5410: Train accuracy = 66.7%
2017-09-09 16:27:35.678131: Step 5410: Cross entropy = 0.570245
2017-09-09 16:27:35.689687: Step 5410: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:35.807579: Step 5420: Train accuracy = 73.3%
2017-09-09 16:27:35.807621: Step 5420: Cross entropy = 0.406773
2017-09-09 16:27:35.819113: Step 5420: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:35.937712: Step 5430: Train accuracy = 80.0%
2017-09-09 16:27:35.937754: Step 5430: Cross entropy = 0.373300
2017-09-09 16:27:35.949230: Step 5430: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:36.067608: Step 5440: Train accuracy = 90.0%
2017-09-09 16:27:36.067650: Step 5440: Cross entropy = 0.365249
2017-09-09 16:27:36.079098: Step 5440: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:36.197643: Step 5450: Train accuracy = 80.0%
2017-09-09 16:27:36.197685: Step 5450: Cross entropy = 0.431398
2017-09-09 16:27:36.209228: Step 5450: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:36.327802: Step 5460: Train accuracy = 70.0%
2017-09-09 16:27:36.327844: Step 5460: Cross entropy = 0.480777
2017-09-09 16:27:36.339295: Step 5460: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:36.457563: Step 5470: Train accuracy = 86.7%
2017-09-09 16:27:36.457604: Step 5470: Cross entropy = 0.351615
2017-09-09 16:27:36.469131: Step 5470: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:36.588141: Step 5480: Train accuracy = 76.7%
2017-09-09 16:27:36.588185: Step 5480: Cross entropy = 0.446797
2017-09-09 16:27:36.599628: Step 5480: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:36.718452: Step 5490: Train accuracy = 86.7%
2017-09-09 16:27:36.718493: Step 5490: Cross entropy = 0.346029
2017-09-09 16:27:36.730832: Step 5490: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:36.850295: Step 5500: Train accuracy = 80.0%
2017-09-09 16:27:36.850336: Step 5500: Cross entropy = 0.452240
2017-09-09 16:27:36.862787: Step 5500: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:36.981753: Step 5510: Train accuracy = 70.0%
2017-09-09 16:27:36.981795: Step 5510: Cross entropy = 0.451751
2017-09-09 16:27:36.993358: Step 5510: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:37.112344: Step 5520: Train accuracy = 86.7%
2017-09-09 16:27:37.112384: Step 5520: Cross entropy = 0.519459
2017-09-09 16:27:37.123850: Step 5520: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:37.241803: Step 5530: Train accuracy = 83.3%
2017-09-09 16:27:37.241845: Step 5530: Cross entropy = 0.366113
2017-09-09 16:27:37.253204: Step 5530: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:37.371825: Step 5540: Train accuracy = 86.7%
2017-09-09 16:27:37.371868: Step 5540: Cross entropy = 0.317330
2017-09-09 16:27:37.383376: Step 5540: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:37.502120: Step 5550: Train accuracy = 86.7%
2017-09-09 16:27:37.502163: Step 5550: Cross entropy = 0.341507
2017-09-09 16:27:37.513652: Step 5550: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:37.631659: Step 5560: Train accuracy = 90.0%
2017-09-09 16:27:37.631702: Step 5560: Cross entropy = 0.263225
2017-09-09 16:27:37.643245: Step 5560: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:37.760657: Step 5570: Train accuracy = 80.0%
2017-09-09 16:27:37.760699: Step 5570: Cross entropy = 0.441186
2017-09-09 16:27:37.773139: Step 5570: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:37.891508: Step 5580: Train accuracy = 90.0%
2017-09-09 16:27:37.891552: Step 5580: Cross entropy = 0.324735
2017-09-09 16:27:37.902931: Step 5580: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:38.027014: Step 5590: Train accuracy = 96.7%
2017-09-09 16:27:38.027061: Step 5590: Cross entropy = 0.248249
2017-09-09 16:27:38.038543: Step 5590: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:38.157774: Step 5600: Train accuracy = 96.7%
2017-09-09 16:27:38.157816: Step 5600: Cross entropy = 0.273698
2017-09-09 16:27:38.169319: Step 5600: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:38.288027: Step 5610: Train accuracy = 90.0%
2017-09-09 16:27:38.288073: Step 5610: Cross entropy = 0.352832
2017-09-09 16:27:38.299476: Step 5610: Validation accuracy = 90.0% (N=30)
2017-09-09 16:27:38.417995: Step 5620: Train accuracy = 83.3%
2017-09-09 16:27:38.418037: Step 5620: Cross entropy = 0.326254
2017-09-09 16:27:38.429544: Step 5620: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:38.547960: Step 5630: Train accuracy = 80.0%
2017-09-09 16:27:38.548002: Step 5630: Cross entropy = 0.414238
2017-09-09 16:27:38.559538: Step 5630: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:38.679055: Step 5640: Train accuracy = 90.0%
2017-09-09 16:27:38.679096: Step 5640: Cross entropy = 0.381434
2017-09-09 16:27:38.690595: Step 5640: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:38.809682: Step 5650: Train accuracy = 93.3%
2017-09-09 16:27:38.809724: Step 5650: Cross entropy = 0.259427
2017-09-09 16:27:38.821334: Step 5650: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:38.940128: Step 5660: Train accuracy = 100.0%
2017-09-09 16:27:38.940170: Step 5660: Cross entropy = 0.231500
2017-09-09 16:27:38.951678: Step 5660: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:39.069351: Step 5670: Train accuracy = 83.3%
2017-09-09 16:27:39.069392: Step 5670: Cross entropy = 0.419553
2017-09-09 16:27:39.080803: Step 5670: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:39.199036: Step 5680: Train accuracy = 70.0%
2017-09-09 16:27:39.199078: Step 5680: Cross entropy = 0.566039
2017-09-09 16:27:39.211157: Step 5680: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:39.329203: Step 5690: Train accuracy = 73.3%
2017-09-09 16:27:39.329244: Step 5690: Cross entropy = 0.385029
2017-09-09 16:27:39.340750: Step 5690: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:39.459630: Step 5700: Train accuracy = 73.3%
2017-09-09 16:27:39.459671: Step 5700: Cross entropy = 0.491191
2017-09-09 16:27:39.471180: Step 5700: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:39.588990: Step 5710: Train accuracy = 83.3%
2017-09-09 16:27:39.589032: Step 5710: Cross entropy = 0.399895
2017-09-09 16:27:39.600459: Step 5710: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:39.719025: Step 5720: Train accuracy = 80.0%
2017-09-09 16:27:39.719069: Step 5720: Cross entropy = 0.447307
2017-09-09 16:27:39.730539: Step 5720: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:39.848393: Step 5730: Train accuracy = 80.0%
2017-09-09 16:27:39.848436: Step 5730: Cross entropy = 0.437177
2017-09-09 16:27:39.860261: Step 5730: Validation accuracy = 63.3% (N=30)
2017-09-09 16:27:39.978792: Step 5740: Train accuracy = 76.7%
2017-09-09 16:27:39.978834: Step 5740: Cross entropy = 0.455666
2017-09-09 16:27:39.990355: Step 5740: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:40.108563: Step 5750: Train accuracy = 86.7%
2017-09-09 16:27:40.108604: Step 5750: Cross entropy = 0.398553
2017-09-09 16:27:40.120304: Step 5750: Validation accuracy = 70.0% (N=30)
2017-09-09 16:27:40.239128: Step 5760: Train accuracy = 83.3%
2017-09-09 16:27:40.239169: Step 5760: Cross entropy = 0.356114
2017-09-09 16:27:40.250747: Step 5760: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:40.371237: Step 5770: Train accuracy = 83.3%
2017-09-09 16:27:40.371282: Step 5770: Cross entropy = 0.385977
2017-09-09 16:27:40.382808: Step 5770: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:40.501926: Step 5780: Train accuracy = 76.7%
2017-09-09 16:27:40.501968: Step 5780: Cross entropy = 0.515888
2017-09-09 16:27:40.513491: Step 5780: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:40.631402: Step 5790: Train accuracy = 83.3%
2017-09-09 16:27:40.631443: Step 5790: Cross entropy = 0.398399
2017-09-09 16:27:40.643147: Step 5790: Validation accuracy = 93.3% (N=30)
2017-09-09 16:27:40.761742: Step 5800: Train accuracy = 76.7%
2017-09-09 16:27:40.761782: Step 5800: Cross entropy = 0.332401
2017-09-09 16:27:40.773340: Step 5800: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:40.892243: Step 5810: Train accuracy = 80.0%
2017-09-09 16:27:40.892285: Step 5810: Cross entropy = 0.343063
2017-09-09 16:27:40.903873: Step 5810: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:41.022727: Step 5820: Train accuracy = 93.3%
2017-09-09 16:27:41.022769: Step 5820: Cross entropy = 0.297864
2017-09-09 16:27:41.034364: Step 5820: Validation accuracy = 73.3% (N=30)
2017-09-09 16:27:41.152948: Step 5830: Train accuracy = 70.0%
2017-09-09 16:27:41.152989: Step 5830: Cross entropy = 0.522374
2017-09-09 16:27:41.164663: Step 5830: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:41.283428: Step 5840: Train accuracy = 96.7%
2017-09-09 16:27:41.283470: Step 5840: Cross entropy = 0.263738
2017-09-09 16:27:41.295018: Step 5840: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:41.412962: Step 5850: Train accuracy = 93.3%
2017-09-09 16:27:41.413003: Step 5850: Cross entropy = 0.325189
2017-09-09 16:27:41.424476: Step 5850: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:41.543447: Step 5860: Train accuracy = 90.0%
2017-09-09 16:27:41.543490: Step 5860: Cross entropy = 0.325270
2017-09-09 16:27:41.555005: Step 5860: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:41.673049: Step 5870: Train accuracy = 76.7%
2017-09-09 16:27:41.673092: Step 5870: Cross entropy = 0.455280
2017-09-09 16:27:41.684559: Step 5870: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:41.802866: Step 5880: Train accuracy = 70.0%
2017-09-09 16:27:41.802910: Step 5880: Cross entropy = 0.481720
2017-09-09 16:27:41.814400: Step 5880: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:41.932651: Step 5890: Train accuracy = 76.7%
2017-09-09 16:27:41.932695: Step 5890: Cross entropy = 0.411545
2017-09-09 16:27:41.944307: Step 5890: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:42.062389: Step 5900: Train accuracy = 80.0%
2017-09-09 16:27:42.062433: Step 5900: Cross entropy = 0.397087
2017-09-09 16:27:42.073912: Step 5900: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:42.192456: Step 5910: Train accuracy = 73.3%
2017-09-09 16:27:42.192497: Step 5910: Cross entropy = 0.506510
2017-09-09 16:27:42.203970: Step 5910: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:42.322466: Step 5920: Train accuracy = 90.0%
2017-09-09 16:27:42.322508: Step 5920: Cross entropy = 0.350994
2017-09-09 16:27:42.334050: Step 5920: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:42.451880: Step 5930: Train accuracy = 86.7%
2017-09-09 16:27:42.451921: Step 5930: Cross entropy = 0.364483
2017-09-09 16:27:42.463337: Step 5930: Validation accuracy = 83.3% (N=30)
2017-09-09 16:27:42.581408: Step 5940: Train accuracy = 90.0%
2017-09-09 16:27:42.581448: Step 5940: Cross entropy = 0.367156
2017-09-09 16:27:42.592851: Step 5940: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:42.711013: Step 5950: Train accuracy = 76.7%
2017-09-09 16:27:42.711056: Step 5950: Cross entropy = 0.371516
2017-09-09 16:27:42.722380: Step 5950: Validation accuracy = 86.7% (N=30)
2017-09-09 16:27:42.841919: Step 5960: Train accuracy = 73.3%
2017-09-09 16:27:42.841960: Step 5960: Cross entropy = 0.426627
2017-09-09 16:27:42.853500: Step 5960: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:42.971704: Step 5970: Train accuracy = 83.3%
2017-09-09 16:27:42.971755: Step 5970: Cross entropy = 0.378455
2017-09-09 16:27:42.983257: Step 5970: Validation accuracy = 66.7% (N=30)
2017-09-09 16:27:43.102142: Step 5980: Train accuracy = 76.7%
2017-09-09 16:27:43.102183: Step 5980: Cross entropy = 0.460707
2017-09-09 16:27:43.113894: Step 5980: Validation accuracy = 76.7% (N=30)
2017-09-09 16:27:43.231609: Step 5990: Train accuracy = 86.7%
2017-09-09 16:27:43.231652: Step 5990: Cross entropy = 0.326965
2017-09-09 16:27:43.243202: Step 5990: Validation accuracy = 80.0% (N=30)
2017-09-09 16:27:43.350670: Step 5999: Train accuracy = 83.3%
2017-09-09 16:27:43.350712: Step 5999: Cross entropy = 0.432477
2017-09-09 16:27:43.362254: Step 5999: Validation accuracy = 76.7% (N=30)
Final test accuracy = 82.0% (N=100)
Converted 2 variables to const ops.

py@py-System-Product-Name:~/deep_learning/TFLearn/re$ python afterselect.py ~/480.jpg
/home/py/480.jpg
2017-09-09 16:35:49.144992: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:35:49.145017: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:35:49.145021: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:35:49.145025: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:35:49.145028: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn’t compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.
2017-09-09 16:35:49.287868: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2017-09-09 16:35:49.288190: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 0 with properties:
name: GeForce GTX 1080
major: 6 minor: 1 memoryClockRate (GHz) 1.8225
pciBusID 0000:01:00.0
Total memory: 7.92GiB
Free memory: 7.49GiB
2017-09-09 16:35:49.384312: W tensorflow/stream_executor/cuda/cuda_driver.cc:523] A non-primary context 0x32df630 exists before initializing the StreamExecutor. We haven’t verified StreamExecutor works with that.
2017-09-09 16:35:49.384572: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2017-09-09 16:35:49.384873: I tensorflow/core/common_runtime/gpu/gpu_device.cc:940] Found device 1 with properties:
name: GeForce GTX 1080
major: 6 minor: 1 memoryClockRate (GHz) 1.8225
pciBusID 0000:02:00.0
Total memory: 7.92GiB
Free memory: 7.81GiB
2017-09-09 16:35:49.385386: I tensorflow/core/common_runtime/gpu/gpu_device.cc:961] DMA: 0 1
2017-09-09 16:35:49.385394: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 0: Y Y
2017-09-09 16:35:49.385399: I tensorflow/core/common_runtime/gpu/gpu_device.cc:971] 1: Y Y
2017-09-09 16:35:49.385405: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1080, pci bus id: 0000:01:00.0)
2017-09-09 16:35:49.385409: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Creating TensorFlow device (/gpu:1) -> (device: 1, name: GeForce GTX 1080, pci bus id: 0000:02:00.0)
2017-09-09 16:35:49.620558: W tensorflow/core/framework/op_def_util.cc:332] Op BatchNormWithGlobalNormalization is deprecated. It will cease to work in GraphDef version 9. Use tf.nn.batch_normalization().
(‘baddp’, 0.81489182)
(‘gooddp’, 0.18510816)