neon.backends.backend.Block

class neon.backends.backend.Block[source]

Bases: object

Simple class that identifies different elements of the computation required to train or run inference on neural networks.

epoch

start of a particular training epoch

minibatch

start processing of a particular mini-batched data partition

fprop

start of forward propagation call for a particular minibatch

bprop

start of backward propagation call for a particular minibatch

update

start of parameter update call for a particular minibatch

__init__()

Initialize self. See help(type(self)) for accurate signature.

bprop = 3
epoch = 0
fprop = 2
minibatch = 1
update = 4