It takes as input a list of tensors, all of the same shape expect for the concatenation axis, and returns a single tensor, the concatenation of all inputs.
layer_concatenate(inputs, axis = -1L)
A list of input tensors (at least 2).
Concatenation axis.
A tensor, the concatenation of the inputs alongside axis axis
.
Other merge layers: layer_add
,
layer_average
, layer_dot
,
layer_maximum
, layer_multiply