小编sha*_*der的帖子

对于在第一个转换层之后的转换层,Tensorflow梯度始终为零.我已经尝试了不同的方法来检查,但渐变总是为零!这是可以运行以检查的小型可重现代码.

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import tensorflow as tf
import numpy as np
import math
import os
import random
import tflearn
batch_size = 100
start = 0
end = batch_size
learning_rate = 0.000001
num_classes = 4
time_steps = 4
embedding = 2
step = 1
_units = 500
num_of_filters = 1000

train_set_x = [[[1,2],[3,4],[5,6],[7,8]],[[1,2],[3,4],[5,6],[7,8]]]
train_set_y = [0,1]

X = tf.placeholder(tf.float32, [None,time_steps,embedding])
Y = tf.placeholder(tf.int32, [None])


x = tf.expand_dims(X,3)

filter_shape = [1, embedding, …