小编roo*_*ish的帖子

我正在尝试构建一个自定义变量自动编码器网络,我在初始化解码器权重时使用来自编码器层的权重转置,我找不到原生的东西,tf.contrib.layers.fully_connected所以我用了tf.assign代替,这里是我的代码层:

def inference_network(inputs, hidden_units, n_outputs):
    """Layer definition for the encoder layer."""
    net = inputs
    with tf.variable_scope('inference_network', reuse=tf.AUTO_REUSE):
        for layer_idx, hidden_dim in enumerate(hidden_units):
            net = layers.fully_connected(
                net,
                num_outputs=hidden_dim,
                weights_regularizer=layers.l2_regularizer(training_params.weight_decay),
                scope='inf_layer_{}'.format(layer_idx))
            add_layer_summary(net)
        z_mean = layers.fully_connected(net, num_outputs=n_outputs, activation_fn=None)
        z_log_sigma = layers.fully_connected(
            net, num_outputs=n_outputs, activation_fn=None)

    return z_mean, z_log_sigma


def generation_network(inputs, decoder_units, n_x):
    """Define the decoder network."""
    net = inputs  # inputs here is the latent representation.
    with tf.variable_scope("generation_network", reuse=tf.AUTO_REUSE):
        assert(len(decoder_units) >= 2)
        # First layer does not have a regularizer …

在下面的简单代码中,当我指定dearness=(40*basic)/100 or rent=(20*basic)/100它时工作正常,但是按照以下方式进行,亲属和租金的赋值语句没有任何效果

请暗示可能的原因.

#include "stdio.h"
#include "conio.h"

int main()
{
    int salary=0,basic=0,dearness=0,rent=0;
    clrscr();
    printf("enter basic salary");
    scanf("%d",&basic);
    dearness=(40/100)*basic; //no effect
    rent=(20/100)*basic;  //no effect
    salary=basic+dearness+rent;
    printf("%d %d",dearness,rent);
    printf("\n%d",salary);
    getch();
    return 0;
}

谢谢.