Question

我是关于tf.train.batch的新手，所以我写了一个样本来测试它。当我运行代码时，我没有得到任何结果，并且该过程仍在运行。

你以前遇到过同样的情况吗？非常感谢提前！

import os

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import numpy as np
import tensorflow as tf


a = [[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]]
b = [1,2,3,4]
input_queue = tf.train.slice_input_producer([a, b],num_epochs=None,shuffle=False)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(sess=sess, coord=coord)
    for i in range(4):


        x,y = tf.train.batch([a,b], batch_size=2)


        x_,y_ =sess.run([x,y])
        print(x_,y_)

    coord.request_stop()
    coord.join(threads)

另外，函数tf.train.slice_input_producer有效。当我忽略tf.train.batch时，代码变为：

import os

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import numpy as np
import tensorflow as tf


a = [[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]]
b = [1,2,3,4]
input_queue = tf.train.slice_input_producer([a, b],num_epochs=None,shuffle=False)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(sess=sess, coord=coord)
    for i in range(4):

     print(sess.run(input_queue))

coord.request_stop()
coord.join(threads)

结果是：

[array([1, 2, 3, 4]), 1]
[array([1, 2, 3, 4]), 2]
[array([1, 2, 3, 4]), 3]
[array([1, 2, 3, 4]), 4]

Answer 1

我认为主要问题是您没有将enqueue_many指定为True，因此它只会重复整批。您可以在official document了解更多信息。这是一个工作样本：

a = [[1,2,3,4],[1,2,3,4],[1,2,3,4],[1,2,3,4]]
b = [1,2,3,4]
a, b = tf.train.batch([a,b], batch_size=1, num_threads=1, capacity=4, enqueue_many=True)
with tf.Session() as sess:
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(sess, coord)
    for i in range(4):
        print(sess.run([a,b]))
    coord.request_stop()
    coord.join(threads)

Tensorflow，tf.train.batch，没有结果

1 个答案: