There are 83 nodes that can be used as predessesor
for a node with an input port of type Keras Deep Learning Network.
An global average pooling 2D layer.
A global max pooling 1D layer.
A global max pooling 2D layer.
A max pooling 1D layer.
A max pooling 2D layer.
A max pooling 3D layer.
2D Convolutional Long-Short Term Memory (LSTM) layer.
2D Convolutional Long-Short Term Memory (LSTM) layer.
Fast GRU implementation backed by CuDNN.
Fast GRU implementation backed by CuDNN.