How to define a layer that can handle different input shape? It looks like this:
# dynamic_input_shape_conv_layer is a tf.keras layer
input_1 (None, 32, 32, 3) => dynamic_input_shape_conv_layer => (None, 32, 32, 6)
input_1 (None, 32, 32, 2) => dynamic_input_shape_conv_layer => (None, 32, 32, 6)
Is that possible to achieve this in TensorFlow 2.x?