for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import theano.tensor as T
from . import NeuralLayer
class Maxout(NeuralLayer):
"""
Maxout activation unit.
- http://arxiv.org/pdf/1302.4389.pdf
def __init__(self, output_dim, num_pieces=4, init=None):
:param num_pieces: pieces of sub maps
super(Maxout, self).__init__("maxout")
self.num_pieces = num_pieces
self.output_dim = output_dim
self.init = init
def prepare(self):
self.W = self.create_weight(shape=(self.num_pieces, self.input_dim, self.output_dim),
initializer=self.init, suffix="maxout")
self.B = self.create_bias(shape=(self.num_pieces, self.output_dim), suffix="maxout")
self.register_parameters(self.W, self.B)
def output(self, x):
return T.max(T.dot(x, self.W) + self.B, axis=1)
uaca /
deepy
