Question

我正在追踪cifar10示例代码here 。我在下面的代码片段中得到了一些问题

def _variable_on_cpu(name, shape, initializer):
 """Helper to create a Variable stored on CPU memory.
      Args:
        name: name of the variable
        shape: list of ints
        initializer: initializer for Variable

      Returns:
        Variable Tensor
      """
      with tf.device('/cpu:0'):
          dtype = tf.float16 if FLAGS.use_fp16 else tf.float32
          var = tf.get_variable(name, shape, initializer=initializer, dtype=dtype)
return var

似乎没有初始化

var（无论是global_variables_initializer（）还是var.initializer.run（））它运行得很好。

但是，以下测试代码会引发错误：

＆＃34; FailedPreconditionError：尝试使用未初始化的值test / x＆＃34;

import tensorflow as tf 
with tf.variable_scope('test') as scope:
    x = tf.get_variable('x', shape = [3, 4], initializer = tf.constant_initializer([0,1,2]))
sess = tf.InteractiveSession()
print(sess.run(x))

有没有人有任何想法？

Answer 1

您在底部添加的代码无效，因为您尚未初始化变量。所以只需添加初始化：

import tensorflow as tf 
with tf.variable_scope('test') as scope:
    x = tf.get_variable('x', shape = [3, 4], initializer = tf.constant_initializer([0,1,2]))
sess = tf.InteractiveSession()
sess.run(tf.global_variables_initializer())
print(sess.run(x))

Tensorflow CIFAR10 get_variable未初始化

1 个答案: