Chinese-Text-Classification
Chinese-Text-Classification copied to clipboard
用新版本中的方法定义输出层,怎么添加l2正则到loss中
self.scores =tf.layers.dense(self.h_drop,num_classes,name='scores',kernel_initializer=tf.contrib.layers.xavier_initializer(),bias_initializer=tf.constant_initializer(0.1))
# W = tf.get_variable(
# "W",
# shape=[num_filters_total, num_classes],
# initializer=tf.contrib.layers.xavier_initializer())
# b = tf.Variable(tf.constant(0.1, shape=[num_classes]), name="b")
# l2_loss += tf.nn.l2_loss(W)
# l2_loss += tf.nn.l2_loss(b)
# print(self.scores.shape)
# self.scores = tf.nn.xw_plus_b(self.h_drop, W, b, name="scores")
self.predictions = tf.argmax(self.scores, 1, name="predictions")
在这里面被注释的这些,我用开头的一句话tf.layers.dense代替了,这个方法有kernel_regularizer,但是不知道这里的kernel_regularizer会不会加入到loss中去。