November 2017
Intermediate to advanced
274 pages
6h 16m
English
Next, we flatten the layer to the num of images and the num of features, which is 16,384 in this case. If you notice for the last layer's output, we have flattened it with the following logic, 16 x 16 x 64 = 16,384:
layer_flat, num_features = flatten_layer(layer_conv3)
If we print these values, you will see the following output:
Tensor("Reshape_1:0", shape=(?, 16384), dtype=float32)16384
Read now
Unlock full access