diff --git a/demos/src/inception-v3-arch.js b/demos/src/inception-v3-arch.js
new file mode 100644
index 0000000..bbdc612
--- /dev/null
+++ b/demos/src/inception-v3-arch.js
@@ -0,0 +1,2120 @@
+export const ARCHITECTURE_DIAGRAM = [
+
+  // /////////////////////////////////////////////////////////////////////
+  // initial
+
+  {
+    name: 'convolution2d_1',
+    className: 'Convolution2D',
+    details: '32 3x3 filters, 2x2 strides, border mode valid, ReLU',
+    row: 0,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_1',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 1,
+    col: 0
+  },
+  {
+    name: 'convolution2d_2',
+    className: 'Convolution2D',
+    details: '32 3x3 filters, 1x1 strides, border mode valid, ReLU',
+    row: 2,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_2',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 3,
+    col: 0
+  },
+  {
+    name: 'convolution2d_3',
+    className: 'Convolution2D',
+    details: '64 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 4,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_3',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 5,
+    col: 0
+  },
+  {
+    name: 'maxpooling2d_1',
+    className: 'MaxPooling2D',
+    details: '3x3 pool size, 2x2 strides, border mode valid',
+    row: 6,
+    col: 0
+  },
+  {
+    name: 'convolution2d_4',
+    className: 'Convolution2D',
+    details: '80 1x1 filters, 1x1 strides, border mode valid, ReLU',
+    row: 7,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_4',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 8,
+    col: 0
+  },
+  {
+    name: 'convolution2d_5',
+    className: 'Convolution2D',
+    details: '192 3x3 filters, 1x1 strides, border mode valid, ReLU',
+    row: 9,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_5',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 10,
+    col: 0
+  },
+  {
+    name: 'maxpooling2d_2',
+    className: 'MaxPooling2D',
+    details: '3x3 pool size, 2x2 strides, border mode valid',
+    row: 11,
+    col: 0
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 0: 35 x 35 x 256
+
+  {
+    name: 'convolution2d_9',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 12,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_9',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 13,
+    col: 0
+  },
+  {
+    name: 'convolution2d_7',
+    className: 'Convolution2D',
+    details: '48 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 12,
+    col: 1
+  },
+  {
+    name: 'convolution2d_10',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 14,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_7',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 13,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_10',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 15,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_1',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 12,
+    col: 3
+  },
+  {
+    name: 'convolution2d_6',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 12,
+    col: 2
+  },
+  {
+    name: 'convolution2d_8',
+    className: 'Convolution2D',
+    details: '64 5x5 filters, 1x1 strides, border mode same, ReLU',
+    row: 14,
+    col: 1
+  },
+  {
+    name: 'convolution2d_11',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 16,
+    col: 2
+  },
+  {
+    name: 'convolution2d_12',
+    className: 'Convolution2D',
+    details: '32 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 13,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_6',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 13,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_8',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 15,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_11',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 17,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_12',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 14,
+    col: 3
+  },
+  {
+    name: 'mixed0',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 18,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 1: 35 x 35 x 256
+
+  {
+    name: 'convolution2d_16',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 19,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_16',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 20,
+    col: 0
+  },
+  {
+    name: 'convolution2d_14',
+    className: 'Convolution2D',
+    details: '48 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 19,
+    col: 1
+  },
+  {
+    name: 'convolution2d_17',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 21,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_14',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 20,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_17',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 22,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_2',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 19,
+    col: 3
+  },
+  {
+    name: 'convolution2d_13',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 19,
+    col: 2
+  },
+  {
+    name: 'convolution2d_15',
+    className: 'Convolution2D',
+    details: '64 5x5 filters, 1x1 strides, border mode same, ReLU',
+    row: 21,
+    col: 1
+  },
+  {
+    name: 'convolution2d_18',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 23,
+    col: 2
+  },
+  {
+    name: 'convolution2d_19',
+    className: 'Convolution2D',
+    details: '32 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 20,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_13',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 20,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_15',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 22,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_18',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 24,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_19',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 21,
+    col: 3
+  },
+  {
+    name: 'mixed1',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 25,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 2: 35 x 35 x 256
+
+  {
+    name: 'convolution2d_23',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 26,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_23',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 27,
+    col: 0
+  },
+  {
+    name: 'convolution2d_21',
+    className: 'Convolution2D',
+    details: '48 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 26,
+    col: 1
+  },
+  {
+    name: 'convolution2d_24',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 28,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_21',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 27,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_24',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 29,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_3',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 26,
+    col: 3
+  },
+  {
+    name: 'convolution2d_20',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 26,
+    col: 2
+  },
+  {
+    name: 'convolution2d_22',
+    className: 'Convolution2D',
+    details: '64 5x5 filters, 1x1 strides, border mode same, ReLU',
+    row: 28,
+    col: 1
+  },
+  {
+    name: 'convolution2d_25',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 30,
+    col: 2
+  },
+  {
+    name: 'convolution2d_26',
+    className: 'Convolution2D',
+    details: '32 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 27,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_20',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 27,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_22',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 29,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_25',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 31,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_26',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 28,
+    col: 3
+  },
+  {
+    name: 'mixed2',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 32,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 3: 17 x 17 x 768
+
+  {
+    name: 'convolution2d_28',
+    className: 'Convolution2D',
+    details: '64 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 33,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_28',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 34,
+    col: 1
+  },
+  {
+    name: 'convolution2d_29',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 35,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_29',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 36,
+    col: 1
+  },
+  {
+    name: 'convolution2d_27',
+    className: 'Convolution2D',
+    details: '384 3x3 filters, 2x2 strides, border mode valid, ReLU',
+    row: 33,
+    col: 0
+  },
+  {
+    name: 'convolution2d_30',
+    className: 'Convolution2D',
+    details: '96 3x3 filters, 2x2 strides, border mode valid, ReLU',
+    row: 37,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_27',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 34,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_30',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 38,
+    col: 1
+  },
+  {
+    name: 'maxpooling2d_3',
+    className: 'MaxPooling2D',
+    details: '3x3 pool size, 2x2 strides, border mode valid',
+    row: 33,
+    col: 2
+  },
+  {
+    name: 'mixed3',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 39,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 4: 17 x 17 x 768
+
+  {
+    name: 'convolution2d_35',
+    className: 'Convolution2D',
+    details: '128 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 40,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_35',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 41,
+    col: 2
+  },
+  {
+    name: 'convolution2d_36',
+    className: 'Convolution2D',
+    details: '128 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 42,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_36',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 43,
+    col: 2
+  },
+  {
+    name: 'convolution2d_32',
+    className: 'Convolution2D',
+    details: '128 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 40,
+    col: 1
+  },
+  {
+    name: 'convolution2d_37',
+    className: 'Convolution2D',
+    details: '128 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 44,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_32',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 41,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_37',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 45,
+    col: 2
+  },
+  {
+    name: 'convolution2d_33',
+    className: 'Convolution2D',
+    details: '128 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 42,
+    col: 1
+  },
+  {
+    name: 'convolution2d_38',
+    className: 'Convolution2D',
+    details: '128 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 46,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_33',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 43,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_38',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 47,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_4',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 40,
+    col: 3
+  },
+  {
+    name: 'convolution2d_31',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 40,
+    col: 0
+  },
+  {
+    name: 'convolution2d_34',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 44,
+    col: 1
+  },
+  {
+    name: 'convolution2d_39',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 48,
+    col: 2
+  },
+  {
+    name: 'convolution2d_40',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 41,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_31',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 41,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_34',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 45,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_39',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 49,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_40',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 42,
+    col: 3
+  },
+  {
+    name: 'mixed4',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 50,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 5: 17 x 17 x 768
+
+  {
+    name: 'convolution2d_45',
+    className: 'Convolution2D',
+    details: '160 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 51,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_45',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 52,
+    col: 2
+  },
+  {
+    name: 'convolution2d_46',
+    className: 'Convolution2D',
+    details: '160 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 53,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_46',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 54,
+    col: 2
+  },
+  {
+    name: 'convolution2d_42',
+    className: 'Convolution2D',
+    details: '160 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 51,
+    col: 1
+  },
+  {
+    name: 'convolution2d_47',
+    className: 'Convolution2D',
+    details: '160 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 55,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_42',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 52,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_47',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 56,
+    col: 2
+  },
+  {
+    name: 'convolution2d_43',
+    className: 'Convolution2D',
+    details: '160 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 53,
+    col: 1
+  },
+  {
+    name: 'convolution2d_48',
+    className: 'Convolution2D',
+    details: '160 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 57,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_43',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 54,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_48',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 58,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_5',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 51,
+    col: 3
+  },
+  {
+    name: 'convolution2d_41',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 51,
+    col: 0
+  },
+  {
+    name: 'convolution2d_44',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 55,
+    col: 1
+  },
+  {
+    name: 'convolution2d_49',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 59,
+    col: 2
+  },
+  {
+    name: 'convolution2d_50',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 52,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_41',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 52,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_44',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 56,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_49',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 60,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_50',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 53,
+    col: 3
+  },
+  {
+    name: 'mixed5',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 61,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 6: 17 x 17 x 768
+
+  {
+    name: 'convolution2d_55',
+    className: 'Convolution2D',
+    details: '160 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 62,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_55',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 63,
+    col: 2
+  },
+  {
+    name: 'convolution2d_56',
+    className: 'Convolution2D',
+    details: '160 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 64,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_56',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 65,
+    col: 2
+  },
+  {
+    name: 'convolution2d_52',
+    className: 'Convolution2D',
+    details: '160 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 62,
+    col: 1
+  },
+  {
+    name: 'convolution2d_57',
+    className: 'Convolution2D',
+    details: '160 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 66,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_52',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 63,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_57',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 67,
+    col: 2
+  },
+  {
+    name: 'convolution2d_53',
+    className: 'Convolution2D',
+    details: '160 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 64,
+    col: 1
+  },
+  {
+    name: 'convolution2d_58',
+    className: 'Convolution2D',
+    details: '160 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 68,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_53',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 65,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_58',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 69,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_6',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 62,
+    col: 3
+  },
+  {
+    name: 'convolution2d_51',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 62,
+    col: 0
+  },
+  {
+    name: 'convolution2d_54',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 66,
+    col: 1
+  },
+  {
+    name: 'convolution2d_59',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 70,
+    col: 2
+  },
+  {
+    name: 'convolution2d_60',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 63,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_51',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 63,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_54',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 67,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_59',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 71,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_60',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 64,
+    col: 3
+  },
+  {
+    name: 'mixed6',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 72,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 7: 17 x 17 x 768
+
+  {
+    name: 'convolution2d_65',
+    className: 'Convolution2D',
+    details: '160 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 73,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_65',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 74,
+    col: 2
+  },
+  {
+    name: 'convolution2d_66',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 75,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_66',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 76,
+    col: 2
+  },
+  {
+    name: 'convolution2d_62',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 73,
+    col: 1
+  },
+  {
+    name: 'convolution2d_67',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 77,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_62',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 74,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_67',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 78,
+    col: 2
+  },
+  {
+    name: 'convolution2d_63',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 75,
+    col: 1
+  },
+  {
+    name: 'convolution2d_68',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 79,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_63',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 76,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_68',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 80,
+    col: 2
+  },
+  {
+    name: 'averagepooling2d_7',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 73,
+    col: 3
+  },
+  {
+    name: 'convolution2d_61',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 73,
+    col: 0
+  },
+  {
+    name: 'convolution2d_64',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 77,
+    col: 1
+  },
+  {
+    name: 'convolution2d_69',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 81,
+    col: 2
+  },
+  {
+    name: 'convolution2d_70',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 74,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_61',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 74,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_64',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 78,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_69',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 82,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_70',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 75,
+    col: 3
+  },
+  {
+    name: 'mixed7',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 83,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 8: 8 x 8 x 1280
+
+  {
+    name: 'convolution2d_73',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 84,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_73',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 85,
+    col: 1
+  },
+  {
+    name: 'convolution2d_74',
+    className: 'Convolution2D',
+    details: '192 1x7 filters, 1x1 strides, border mode same, ReLU',
+    row: 86,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_74',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 87,
+    col: 1
+  },
+  {
+    name: 'convolution2d_71',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 84,
+    col: 0
+  },
+  {
+    name: 'convolution2d_75',
+    className: 'Convolution2D',
+    details: '192 7x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 88,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_71',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 85,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_75',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 89,
+    col: 1
+  },
+  {
+    name: 'convolution2d_72',
+    className: 'Convolution2D',
+    details: '320 3x3 filters, 2x2 strides, border mode valid, ReLU',
+    row: 86,
+    col: 0
+  },
+  {
+    name: 'convolution2d_76',
+    className: 'Convolution2D',
+    details: '192 3x3 filters, 2x2 strides, border mode valid, ReLU',
+    row: 90,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_72',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 87,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_76',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 91,
+    col: 1
+  },
+  {
+    name: 'averagepooling2d_8',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 2x2 strides, border mode valid',
+    row: 84,
+    col: 2
+  },
+  {
+    name: 'mixed8',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 92,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 9: 8 x 8 x 2048
+
+  {
+    name: 'convolution2d_81',
+    className: 'Convolution2D',
+    details: '448 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 93,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_81',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 94,
+    col: 2
+  },
+  {
+    name: 'convolution2d_78',
+    className: 'Convolution2D',
+    details: '384 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 93,
+    col: 1
+  },
+  {
+    name: 'convolution2d_82',
+    className: 'Convolution2D',
+    details: '384 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 95,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_78',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 94,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_82',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 96,
+    col: 2
+  },
+  {
+    name: 'convolution2d_79',
+    className: 'Convolution2D',
+    details: '384 1x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 95,
+    col: 1,
+    subcol: 0
+  },
+  {
+    name: 'convolution2d_80',
+    className: 'Convolution2D',
+    details: '384 3x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 95,
+    col: 1,
+    subcol: 1
+  },
+  {
+    name: 'convolution2d_83',
+    className: 'Convolution2D',
+    details: '384 1x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 97,
+    col: 2,
+    subcol: 0
+  },
+  {
+    name: 'convolution2d_84',
+    className: 'Convolution2D',
+    details: '384 3x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 97,
+    col: 2,
+    subcol: 1
+  },
+  {
+    name: 'averagepooling2d_9',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 93,
+    col: 3
+  },
+  {
+    name: 'convolution2d_77',
+    className: 'Convolution2D',
+    details: '320 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 93,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_79',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 96,
+    col: 1,
+    subcol: 0
+  },
+  {
+    name: 'batchnormalization_80',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 96,
+    col: 1,
+    subcol: 1
+  },
+  {
+    name: 'batchnormalization_83',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 98,
+    col: 2,
+    subcol: 0
+  },
+  {
+    name: 'batchnormalization_84',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 98,
+    col: 2,
+    subcol: 1
+  },
+  {
+    name: 'convolution2d_85',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 94,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_77',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 94,
+    col: 0
+  },
+  {
+    name: 'mixed9_0',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 97,
+    col: 1
+  },
+  {
+    name: 'merge_1',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 99,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_85',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 95,
+    col: 3
+  },
+  {
+    name: 'mixed9',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 100,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // mixed block 10: 8 x 8 x 2048
+
+  {
+    name: 'convolution2d_90',
+    className: 'Convolution2D',
+    details: '448 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 101,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_90',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 102,
+    col: 2
+  },
+  {
+    name: 'convolution2d_87',
+    className: 'Convolution2D',
+    details: '384 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 101,
+    col: 1
+  },
+  {
+    name: 'convolution2d_91',
+    className: 'Convolution2D',
+    details: '384 3x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 103,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_87',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 102,
+    col: 1
+  },
+  {
+    name: 'batchnormalization_91',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 104,
+    col: 2
+  },
+  {
+    name: 'convolution2d_88',
+    className: 'Convolution2D',
+    details: '384 1x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 103,
+    col: 1,
+    subcol: 0
+  },
+  {
+    name: 'convolution2d_89',
+    className: 'Convolution2D',
+    details: '384 3x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 103,
+    col: 1,
+    subcol: 1
+  },
+  {
+    name: 'convolution2d_92',
+    className: 'Convolution2D',
+    details: '384 1x3 filters, 1x1 strides, border mode same, ReLU',
+    row: 105,
+    col: 2,
+    subcol: 0
+  },
+  {
+    name: 'convolution2d_93',
+    className: 'Convolution2D',
+    details: '384 3x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 105,
+    col: 2,
+    subcol: 1
+  },
+  {
+    name: 'averagepooling2d_10',
+    className: 'AveragePooling2D',
+    details: '3x3 pool size, 1x1 strides, border mode same',
+    row: 101,
+    col: 3
+  },
+  {
+    name: 'convolution2d_86',
+    className: 'Convolution2D',
+    details: '320 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 101,
+    col: 0
+  },
+  {
+    name: 'batchnormalization_88',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 104,
+    col: 1,
+    subcol: 0
+  },
+  {
+    name: 'batchnormalization_89',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 104,
+    col: 1,
+    subcol: 1
+  },
+  {
+    name: 'batchnormalization_92',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 106,
+    col: 2,
+    subcol: 0
+  },
+  {
+    name: 'batchnormalization_93',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 106,
+    col: 2,
+    subcol: 1
+  },
+  {
+    name: 'convolution2d_94',
+    className: 'Convolution2D',
+    details: '192 1x1 filters, 1x1 strides, border mode same, ReLU',
+    row: 102,
+    col: 3
+  },
+  {
+    name: 'batchnormalization_86',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 102,
+    col: 0
+  },
+  {
+    name: 'mixed9_1',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 105,
+    col: 1
+  },
+  {
+    name: 'merge_2',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 107,
+    col: 2
+  },
+  {
+    name: 'batchnormalization_94',
+    className: 'BatchNormalization',
+    details: 'channel axis features',
+    row: 103,
+    col: 3
+  },
+  {
+    name: 'mixed10',
+    className: 'Merge',
+    details: 'concat by channel axes',
+    row: 108,
+    col: 3
+  },
+
+  // /////////////////////////////////////////////////////////////////////
+  // final
+
+  {
+    name: 'avg_pool',
+    className: 'AveragePooling2D',
+    details: '8x8 pool size, 8x8 strides, border mode valid',
+    row: 109,
+    col: 3
+  },
+  {
+    name: 'flatten',
+    className: 'Flatten',
+    details: '',
+    row: 110,
+    col: 3
+  },
+  {
+    name: 'predictions',
+    className: 'Dense',
+    details: 'output dimensions 1000, softmax activation',
+    row: 111,
+    col: 3
+  }
+
+]
+
+export const ARCHITECTURE_CONNECTIONS = [
+
+  // main
+
+  {
+    from: 'averagepooling2d_1',
+    to: 'predictions'
+  },
+  {
+    from: 'convolution2d_1',
+    to: 'maxpooling2d_2'
+  },
+
+  // mixed block 0: 35 x 35 x 256
+
+  {
+    from: 'maxpooling2d_2',
+    to: 'batchnormalization_9'
+  },
+  {
+    from: 'maxpooling2d_2',
+    to: 'batchnormalization_11',
+    corner: 'top-right'
+  },
+  {
+    from: 'maxpooling2d_2',
+    to: 'batchnormalization_8',
+    corner: 'top-right'
+  },
+  {
+    from: 'maxpooling2d_2',
+    to: 'averagepooling2d_1',
+    corner: 'top-right'
+  },
+  {
+    from: 'batchnormalization_9',
+    to: 'mixed0',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'batchnormalization_11',
+    to: 'mixed0',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'batchnormalization_8',
+    to: 'mixed0',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 1: 35 x 35 x 256
+
+  {
+    from: 'mixed0',
+    to: 'convolution2d_16',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed0',
+    to: 'convolution2d_14',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed0',
+    to: 'convolution2d_13',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_16',
+    to: 'mixed1',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_14',
+    to: 'mixed1',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_13',
+    to: 'mixed1',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 2: 35 x 35 x 256
+
+  {
+    from: 'mixed1',
+    to: 'convolution2d_23',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed1',
+    to: 'convolution2d_21',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed1',
+    to: 'convolution2d_20',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_23',
+    to: 'mixed2',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_21',
+    to: 'mixed2',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_20',
+    to: 'mixed2',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 3: 17 x 17 x 768
+
+  {
+    from: 'mixed2',
+    to: 'convolution2d_27',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed2',
+    to: 'convolution2d_28',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed2',
+    to: 'maxpooling2d_3',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_27',
+    to: 'mixed3',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_28',
+    to: 'mixed3',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'maxpooling2d_3',
+    to: 'mixed3',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 4: 17 x 17 x 768
+
+  {
+    from: 'mixed3',
+    to: 'convolution2d_31',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed3',
+    to: 'convolution2d_32',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed3',
+    to: 'convolution2d_35',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_31',
+    to: 'mixed4',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_32',
+    to: 'mixed4',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_35',
+    to: 'mixed4',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 5: 17 x 17 x 768
+
+  {
+    from: 'mixed4',
+    to: 'convolution2d_41',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed4',
+    to: 'convolution2d_42',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed4',
+    to: 'convolution2d_45',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_41',
+    to: 'mixed5',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_42',
+    to: 'mixed5',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_45',
+    to: 'mixed5',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 6: 17 x 17 x 768
+
+  {
+    from: 'mixed5',
+    to: 'convolution2d_51',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed5',
+    to: 'convolution2d_52',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed5',
+    to: 'convolution2d_55',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_51',
+    to: 'mixed6',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_52',
+    to: 'mixed6',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_55',
+    to: 'mixed6',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 7: 17 x 17 x 768
+
+  {
+    from: 'mixed6',
+    to: 'convolution2d_61',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed6',
+    to: 'convolution2d_62',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed6',
+    to: 'convolution2d_65',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_61',
+    to: 'mixed7',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_62',
+    to: 'mixed7',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_65',
+    to: 'mixed7',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 8: 8 x 8 x 1280
+
+  {
+    from: 'mixed7',
+    to: 'convolution2d_71',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed7',
+    to: 'convolution2d_73',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed7',
+    to: 'averagepooling2d_8',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'convolution2d_71',
+    to: 'mixed8',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'convolution2d_73',
+    to: 'mixed8',
+    corner: 'bottom-left'
+  },
+  {
+    from: 'averagepooling2d_8',
+    to: 'mixed8',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 9: 8 x 8 x 2048
+
+  {
+    from: 'mixed8',
+    to: 'convolution2d_77',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed8',
+    to: 'convolution2d_78',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed8',
+    to: 'convolution2d_81',
+    corner: 'bottom-right'
+  },
+
+  {
+    from: 'convolution2d_77',
+    to: 'mixed9',
+    corner: 'bottom-left'
+  },
+
+  {
+    from: 'convolution2d_78',
+    to: 'batchnormalization_78'
+  },
+  {
+    from: 'batchnormalization_78',
+    to: 'convolution2d_79'
+  },
+  {
+    from: 'batchnormalization_78',
+    to: 'convolution2d_80'
+  },
+  {
+    from: 'convolution2d_79',
+    to: 'batchnormalization_79'
+  },
+  {
+    from: 'convolution2d_80',
+    to: 'batchnormalization_80'
+  },
+  {
+    from: 'batchnormalization_79',
+    to: 'mixed9_0'
+  },
+  {
+    from: 'batchnormalization_80',
+    to: 'mixed9_0'
+  },
+  {
+    from: 'mixed9_0',
+    to: 'mixed9',
+    corner: 'bottom-left'
+  },
+
+  {
+    from: 'convolution2d_81',
+    to: 'batchnormalization_81'
+  },
+  {
+    from: 'batchnormalization_81',
+    to: 'convolution2d_82'
+  },
+  {
+    from: 'convolution2d_82',
+    to: 'batchnormalization_82'
+  },
+  {
+    from: 'batchnormalization_82',
+    to: 'convolution2d_83'
+  },
+  {
+    from: 'batchnormalization_82',
+    to: 'convolution2d_84'
+  },
+  {
+    from: 'convolution2d_83',
+    to: 'batchnormalization_83'
+  },
+  {
+    from: 'convolution2d_84',
+    to: 'batchnormalization_84'
+  },
+  {
+    from: 'batchnormalization_83',
+    to: 'merge_1'
+  },
+  {
+    from: 'batchnormalization_84',
+    to: 'merge_1'
+  },
+  {
+    from: 'merge_1',
+    to: 'mixed9',
+    corner: 'bottom-left'
+  },
+
+  // mixed block 10: 8 x 8 x 2048
+
+  {
+    from: 'mixed9',
+    to: 'convolution2d_86',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed9',
+    to: 'convolution2d_87',
+    corner: 'bottom-right'
+  },
+  {
+    from: 'mixed9',
+    to: 'convolution2d_90',
+    corner: 'bottom-right'
+  },
+
+  {
+    from: 'convolution2d_86',
+    to: 'mixed10',
+    corner: 'bottom-left'
+  },
+
+  {
+    from: 'convolution2d_87',
+    to: 'batchnormalization_87'
+  },
+  {
+    from: 'batchnormalization_87',
+    to: 'convolution2d_88'
+  },
+  {
+    from: 'batchnormalization_87',
+    to: 'convolution2d_89'
+  },
+  {
+    from: 'convolution2d_88',
+    to: 'batchnormalization_88'
+  },
+  {
+    from: 'convolution2d_89',
+    to: 'batchnormalization_89'
+  },
+  {
+    from: 'batchnormalization_88',
+    to: 'mixed9_1'
+  },
+  {
+    from: 'batchnormalization_89',
+    to: 'mixed9_1'
+  },
+  {
+    from: 'mixed9_1',
+    to: 'mixed10',
+    corner: 'bottom-left'
+  },
+
+  {
+    from: 'convolution2d_90',
+    to: 'batchnormalization_90'
+  },
+  {
+    from: 'batchnormalization_90',
+    to: 'convolution2d_91'
+  },
+  {
+    from: 'convolution2d_91',
+    to: 'batchnormalization_91'
+  },
+  {
+    from: 'batchnormalization_91',
+    to: 'convolution2d_92'
+  },
+  {
+    from: 'batchnormalization_91',
+    to: 'convolution2d_93'
+  },
+  {
+    from: 'convolution2d_92',
+    to: 'batchnormalization_92'
+  },
+  {
+    from: 'convolution2d_93',
+    to: 'batchnormalization_93'
+  },
+  {
+    from: 'batchnormalization_92',
+    to: 'merge_2'
+  },
+  {
+    from: 'batchnormalization_93',
+    to: 'merge_2'
+  },
+  {
+    from: 'merge_2',
+    to: 'mixed10',
+    corner: 'bottom-left'
+  }
+
+]
diff --git a/demos/src/inception-v3.css b/demos/src/inception-v3.css
new file mode 100644
index 0000000..f56af94
--- /dev/null
+++ b/demos/src/inception-v3.css
@@ -0,0 +1,216 @@
+@import './_variables.css';
+
+.demo.inception-v3 {
+  .top-container {
+    margin: 10px;
+    position: relative;
+    display: flex;
+
+    .input-container {
+      .input-label {
+        font-family: $font-3;
+        font-size: 16px;
+        color: $color-2;
+        text-align: left;
+        user-select: none;
+        cursor: default;
+      }
+
+      .image-url {
+        display: flex;
+        flex-direction: row;
+        align-items: center;
+        justify-content: flex-start;
+        position: relative;
+
+        span {
+          margin: 0 10px;
+          font-family: $font-3;
+          font-size: 16px;
+          color: $color-2;
+        }
+      }
+    }
+
+    .controls {
+      width: 250px;
+      margin-left: 40px;
+      display: flex;
+      flex-direction: column;
+      align-items: center;
+      justify-content: center;
+
+      & > .mdl-switch {
+        margin-bottom: 5px;
+
+        .mdl-switch__label {
+          font-size: 12px !important;
+        }
+      }
+    }
+  }
+
+  .columns.input-output {
+    max-width: 800px;
+    margin: 0 auto;
+
+    .column {
+      display: flex;
+      align-items: center;
+      justify-content: center;
+    }
+
+    .column.input-column {
+      position: relative;
+
+      .loading-indicator {
+        position: absolute;
+        top: 0;
+        left: -10px;
+        display: flex;
+        flex-direction: column;
+        align-self: flex-start;
+
+        .mdl-spinner {
+          margin: 20px;
+          align-self: center;
+        }
+
+        .error {
+          color: $color-err;
+          font-size: 14px;
+          font-family: $font-1;
+          margin: 20px;
+        }
+      }
+
+      .canvas-container {
+        display: inline-flex;
+        justify-content: flex-end;
+
+        canvas {
+          background: white;
+        }
+      }
+    }
+
+    .column.output-column {
+      .output {
+        width: 370px;
+        height: 160px;
+        display: flex;
+        flex-direction: column;
+        align-items: flex-start;
+        justify-content: center;
+
+        .output-class {
+          display: flex;
+          flex-direction: row;
+          align-items: center;
+          justify-content: center;
+          padding: 6px 0;
+
+          .output-label {
+            text-align: right;
+            width: 200px;
+            white-space: nowrap;
+            overflow: hidden;
+            text-overflow: ellipsis;
+            font-family: $font-2;
+            font-size: 18px;
+            color: $color-2;
+            padding: 0 6px;
+            border-right: 2px solid $color-1-lighter;
+          }
+
+          .output-bar {
+            height: 8px;
+            transition: width 0.2s ease-out;
+          }
+
+          .output-value {
+            text-align: left;
+            margin-left: 5px;
+            font-family: $font-2;
+            font-size: 14px;
+            color: $color-2;
+          }
+        }
+
+        .output-class.predicted {
+          .output-label {
+            color: $color-1;
+            border-left-color: $color-1;
+          }
+
+          .output-value {
+            color: $color-1;
+          }
+        }
+      }
+    }
+  }
+
+  .architecture-container {
+    min-width: 800px;
+    max-width: 1200px;
+    margin: 0 auto;
+    position: relative;
+
+    .layers-row {
+      display: flex;
+      flex-direction: row;
+      align-items: center;
+      justify-content: center;
+      margin-bottom: 5px;
+      position: relative;
+      z-index: 1;
+
+      .layer-column {
+        flex: 1;
+        display: flex;
+        align-items: center;
+        justify-content: center;
+        padding: 5px;
+
+        .layer {
+          display: inline-block;
+          background: white;
+          border: 2px solid white;
+          border-radius: 5px;
+          padding: 2px 10px 0px;
+          margin: 3px;
+
+          .layer-class-name {
+            color: $color-1;
+            font-size: 14px;
+            font-weight: bold;
+          }
+
+          .layer-details {
+            color: #999999;
+            font-size: 12px;
+            font-weight: bold;
+          }
+        }
+
+        .layer.has-result {
+          border-color: $color-1;
+        }
+      }
+    }
+
+    .architecture-connections {
+      position: absolute;
+      top: 0;
+      left: 0;
+      z-index: 0;
+
+      path {
+        stroke-width: 4px;
+        stroke: #AAAAAA;
+        fill: none;
+      }
+    }
+  }
+}
diff --git a/demos/src/inception-v3.js b/demos/src/inception-v3.js
new file mode 100644
index 0000000..15f6871
--- /dev/null
+++ b/demos/src/inception-v3.js
@@ -0,0 +1,220 @@
+/* global Vue, loadImage */
+import './inception-v3.css'
+
+import ndarray from 'ndarray'
+import ops from 'ndarray-ops'
+import filter from 'lodash/filter'
+import * as utils from './utils'
+import { IMAGE_URLS } from './image-urls'
+import { ARCHITECTURE_DIAGRAM, ARCHITECTURE_CONNECTIONS } from './inception-v3-arch'
+
+const MODEL_FILEPATHS_DEV = {
+  model: '/demos/data/inception_v3/inception_v3.json',
+  weights: '/demos/data/inception_v3/inception_v3_weights.buf',
+  metadata: '/demos/data/inception_v3/inception_v3_metadata.json'
+}
+const MODEL_FILEPATHS_PROD = {
+  model: 'demos/data/inception_v3/inception_v3.json',
+  weights: 'https://transcranial.github.io/keras-js-demos-data/inception_v3/inception_v3_weights.buf',
+  metadata: 'demos/data/inception_v3/inception_v3_metadata.json'
+}
+const MODEL_CONFIG = {
+  filepaths: (process.env.NODE_ENV === 'production') ? MODEL_FILEPATHS_PROD : MODEL_FILEPATHS_DEV
+}
+
+/**
+ *
+ * VUE COMPONENT
+ *
+ */
+export const InceptionV3 = Vue.extend({
+  props: ['hasWebgl'],
+
+  template: require('raw!./inception-v3.template.html'),
+
+  data: function () {
+    return {
+      model: new KerasJS.Model(Object.assign({ gpu: this.hasWebgl, layerCallPauses: true }, MODEL_CONFIG)),
+      modelLoading: true,
+      modelRunning: false,
+      imageURLInput: null,
+      imageURLSelect: null,
+      imageURLSelectList: IMAGE_URLS,
+      imageLoading: false,
+      imageLoadingError: false,
+      output: null,
+      architectureDiagram: ARCHITECTURE_DIAGRAM,
+      architectureConnections: ARCHITECTURE_CONNECTIONS,
+      architectureDiagramPaths: [],
+      useGpu: this.hasWebgl,
+      showComputationFlow: true
+    }
+  },
+
+  computed: {
+    loadingProgress: function () {
+      return this.model.getLoadingProgress()
+    },
+    architectureDiagramRows: function () {
+      let rows = []
+      for (let row = 0; row < 112; row++) {
+        let cols = []
+        for (let col = 0; col < 4; col++) {
+          cols.push(filter(this.architectureDiagram, { row, col }))
+        }
+        rows.push(cols)
+      }
+      return rows
+    },
+    layersWithResults: function () {
+      // store as computed property for reactivity
+      return this.model.layersWithResults
+    },
+    outputClasses: function () {
+      if (!this.output) {
+        let empty = []
+        for (let i = 0; i < 5; i++) {
+          empty.push({ name: '-', probability: 0 })
+        }
+        return empty
+      }
+      return utils.imagenetClassesTopK(this.output, 5)
+    }
+  },
+
+  ready: function () {
+    this.model.ready().then(() => {
+      this.modelLoading = false
+    })
+
+    this.architectureDiagramPaths = []
+    setTimeout(() => {
+      this.architectureConnections.forEach(conn => {
+        const containerElem = document.getElementsByClassName('architecture-container')[0]
+        const fromElem = document.getElementById(conn.from)
+        const toElem = document.getElementById(conn.to)
+        const containerElemCoords = containerElem.getBoundingClientRect()
+        const fromElemCoords = fromElem.getBoundingClientRect()
+        const toElemCoords = toElem.getBoundingClientRect()
+        const xContainer = containerElemCoords.left
+        const yContainer = containerElemCoords.top
+        const xFrom = fromElemCoords.left + fromElemCoords.width / 2 - xContainer
+        const yFrom = fromElemCoords.top + fromElemCoords.height / 2 - yContainer
+        const xTo = toElemCoords.left + toElemCoords.width / 2 - xContainer
+        const yTo = toElemCoords.top + toElemCoords.height / 2 - yContainer
+
+        let path = `M${xFrom},${yFrom} L${xTo},${yTo}`
+        if (conn.corner === 'top-right') {
+          path = `M${xFrom},${yFrom} L${xTo - 10},${yFrom} Q${xTo},${yFrom} ${xTo},${yFrom + 10} L${xTo},${yTo}`
+        } else if (conn.corner === 'bottom-left') {
+          path = `M${xFrom},${yFrom} L${xFrom},${yTo - 10} Q${xFrom},${yTo} ${xFrom + 10},${yTo} L${xTo},${yTo}`
+        } else if (conn.corner === 'top-left') {
+          path = `M${xFrom},${yFrom} L${xTo + 10},${yFrom} Q${xTo},${yFrom} ${xTo},${yFrom + 10} L${xTo},${yTo}`
+        } else if (conn.corner === 'bottom-right') {
+          path = `M${xFrom},${yFrom} L${xFrom},${yFrom + 20} Q${xFrom},${yFrom + 30} ${xFrom - 10},${yFrom + 30} L${xTo + 10},${yFrom + 30} Q${xTo},${yFrom + 30} ${xTo},${yFrom + 40} L${xTo},${yTo}`
+        }
+
+        this.architectureDiagramPaths.push(path)
+      })
+    }, 1000)
+  },
+
+  methods: {
+
+    toggleGpu: function () {
+      this.model.toggleGpu(!this.useGpu)
+    },
+
+    toggleComputationFlow: function () {
+      this.model.layerCallPauses = !this.showComputationFlow
+    },
+
+    imageURLInputChanged: function (e) {
+      this.imageURLSelect = null
+      this.loadImageToCanvas(this.imageURLInput)
+    },
+
+    imageURLSelectChanged: function (e) {
+      this.imageURLInput = this.imageURLSelect
+      this.loadImageToCanvas(this.imageURLSelect)
+    },
+
+    loadImageToCanvas: function (url) {
+      if (!url) {
+        this.clearAll()
+        return
+      }
+
+      this.imageLoading = true
+      loadImage(
+        url,
+        img => {
+          if (img.type === 'error') {
+            this.imageLoadingError = true
+            this.imageLoading = false
+          } else {
+            // load image data onto input canvas
+            const ctx = document.getElementById('input-canvas').getContext('2d')
+            ctx.drawImage(img, 0, 0)
+            this.imageLoadingError = false
+            this.imageLoading = false
+            this.modelRunning = true
+            // model predict
+            this.$nextTick(function () {
+              setTimeout(() => {
+                this.runModel()
+              }, 200)
+            })
+          }
+        },
+        {
+          maxWidth: 299,
+          maxHeight: 299,
+          cover: true,
+          crop: true,
+          canvas: true,
+          crossOrigin: 'Anonymous'
+        }
+      )
+    },
+
+    runModel: function () {
+      const ctx = document.getElementById('input-canvas').getContext('2d')
+      const imageData = ctx.getImageData(0, 0, ctx.canvas.width, ctx.canvas.height)
+      const { data, width, height } = imageData
+
+      // data processing
+      // see https://github.com/fchollet/keras/blob/master/keras/applications/imagenet_utils.py
+      let dataTensor = ndarray(new Float32Array(data), [width, height, 4])
+      let dataProcessedTensor = ndarray(new Float32Array(width * height * 3), [width, height, 3])
+      ops.divseq(dataTensor, 255)
+      ops.subseq(dataTensor, 0.5)
+      ops.mulseq(dataTensor, 2)
+      ops.assign(dataProcessedTensor.pick(null, null, 0), dataTensor.pick(null, null, 0))
+      ops.assign(dataProcessedTensor.pick(null, null, 1), dataTensor.pick(null, null, 1))
+      ops.assign(dataProcessedTensor.pick(null, null, 2), dataTensor.pick(null, null, 2))
+
+      const inputData = {
+        'input_1': dataProcessedTensor.data
+      }
+      this.model.predict(inputData).then(outputData => {
+        this.output = outputData['predictions']
+        this.modelRunning = false
+      })
+    },
+
+    clearAll: function () {
+      this.modelRunning = false
+      this.imageURLInput = null
+      this.imageURLSelect = null
+      this.imageLoading = false
+      this.imageLoadingError = false
+      this.output = null
+
+      this.model.layersWithResults = []
+
+      const ctx = document.getElementById('input-canvas').getContext('2d')
+      ctx.clearRect(0, 0, ctx.canvas.width, ctx.canvas.height)
+    }
+  }
+})
diff --git a/demos/src/inception-v3.template.html b/demos/src/inception-v3.template.html
new file mode 100644
index 0000000..c63f7bf
--- /dev/null
+++ b/demos/src/inception-v3.template.html
@@ -0,0 +1,97 @@
+<div class="demo inception-v3">
+  <div class="title">
+    <span>Inception V3, trained on ImageNet</span>
+    <mdl-spinner v-if="modelLoading && loadingProgress < 100"></mdl-spinner>
+  </div>
+  <div class="loading-progress" v-if="modelLoading && loadingProgress < 100">
+    Loading...{{ loadingProgress }}%
+  </div>
+  <div class="info-panel">
+    <div class="info-panel-text">
+      <a href="https://www.flickr.com/search/?text=&license=2%2C3%2C4%2C5%2C6%2C9&sort=interestingness-desc">Flickr creative commons</a>.
+    </div>
+    <div class="info-panel-close">
+      <i class="material-icons">close</i>CLOSE
+    </div>
+  </div>
+  <div class="top-container" v-if="!modelLoading">
+    <div class="input-container">
+      <div class="input-label">Enter a valid image URL or select an image from the dropdown:</div>
+      <div class="image-url">
+        <mdl-textfield
+          floating-label="enter image url"
+          :value.sync="imageURLInput"
+          spellcheck="false"
+          @change="imageURLInputChanged"
+        ></mdl-textfield>
+        <span>or</span>
+        <mdl-select
+          label="select image"
+          id="image-url-select"
+          :value.sync="imageURLSelect"
+          :options="imageURLSelectList"
+          style="width:200px;"
+          @change="imageURLSelectChanged"
+        ></mdl-select>
+      </div>
+    </div>
+    <div class="controls">
+      <mdl-switch
+        :checked.sync="useGpu"
+        :disabled="modelLoading || modelRunning || !hasWebgl"
+        @click="toggleGpu"
+      >Use GPU</mdl-switch>
+      <mdl-switch
+        :checked.sync="showComputationFlow"
+        :disabled="modelLoading || modelRunning"
+        @click="toggleComputationFlow"
+      >Show computation flow</mdl-switch>
+    </div>
+  </div>
+  <div class="columns input-output" v-if="!modelLoading">
+    <div class="column input-column">
+      <div class="loading-indicator">
+        <mdl-spinner v-if="imageLoading || modelRunning"></mdl-spinner>
+        <div class="error" v-if="imageLoadingError">Error loading URL</div>
+      </div>
+      <div class="canvas-container">
+        <canvas id="input-canvas" width="299" height="299"></canvas>
+      </div>
+    </div>
+    <div class="column output-column">
+      <div class="output">
+        <div class="output-class"
+          :class="{ 'predicted': i === 0 && outputClasses[i].probability.toFixed(2) > 0 }"
+          v-for="i in [0, 1, 2, 3, 4]"
+        >
+          <div class="output-label">{{ outputClasses[i].name }}</div>
+          <div class="output-bar"
+            style="width: {{ Math.round(100 * outputClasses[i].probability) }}px; background: rgba(27, 188, 155, {{ outputClasses[i].probability.toFixed(2) }});"
+          ></div>
+          <div class="output-value">{{ Math.round(100 * outputClasses[i].probability) }}%</div>
+        </div>
+      </div>
+    </div>
+  </div>
+  <div class="architecture-container" v-if="!modelLoading">
+    <div v-for="row in architectureDiagramRows" class="layers-row">
+      <div v-for="layers in row" class="layer-column">
+        <div
+          v-for="layer in layers"
+          v-if="layer.className"
+          class="layer"
+          :class="{ 'has-result': layersWithResults.includes(layer.name) }"
+          :id="layer.name"
+        >
+          <div class="layer-class-name">{{ layer.className }}</div>
+          <div class="layer-details"> {{ layer.details }}</div>
+        </div>
+      </div>
+    </div>
+    <svg class="architecture-connections" width="100%" height="100%">
+      <g>
+        <path v-for="path in architectureDiagramPaths" :d="path" />
+      </g>
+    </svg>
+  </div>
+</div>