keraflow.layers.core.Reshape Class Reference

Reshapes the input tensor according to a given pattern. More...

Inheritance diagram for keraflow.layers.core.Reshape:

Public Member Functions
def	__init__

Detailed Description

Reshapes the input tensor according to a given pattern.

Behaves like numpy.reshape().

• input_shape: nD, (nb_samples, x, y, ...)
• output_shape: ?D, (nb_samples, a*y, ...)

Constructor & Destructor Documentation

def keraflow.layers.core.Reshape.__init__	(		self,
			target_shape,
			include_batch_dim = False,
			kwargs
	)

Parameters

target_shape	list of int. Target shape. You could specify at most one -1, which will be automatically determined.
include_batch_dim	boolean. When False, in target_shape, 0 refers to the second dimension of the tensor, i.e. the dimension next to the batch size dimension. Otherwise, 0 refers to the first (batch) dimension.
kwargs	see Layer.__init__.