TensorFlow 报错 TypeError Fetch argument 0.16 has invalid type class 'numpy.float32', must be a string

报错代码：

with tf.Session() as sess:
    sess.run(init_op)

    for i in range(self.epoch_num):
        batch_images, batch_labels = mnist.train.next_batch(self.batch_size)
        batch_images = tf.reshape(tensor=batch_images, shape=[self.batch_size, 28, 28, 1])
        batch_images = tf.image.resize_images(images=batch_images,size=(32,32))
        print("images shape:{}".format(batch_images.shape))
        print("labels shape:{}".format(batch_labels.shape))
        sess.run(train_op, feed_dict={images_holder:batch_images.eval(), labels_holder:batch_labels})
        accuracy = sess.run(fetches=accuracy, feed_dict={images_holder: batch_images.eval(), labels_holder: batch_labels})
        print(accuracy)

报错信息：
TypeError: Fetch argument 0.16 has invalid type <class 'numpy.float32'>, must be a string or Tensor. (Can not convert a float32 into a Tensor or Operation.)
 

仔细检查代码后，问题主要出在这一句上：accuracy = sess.run(accuracy, feed_dict={images_holder: batch_images.eval(), labels_holder: batch_labels})

 

解决方法：将接受返回值的变量 accuracy 换一个名字，与 sess.run() 中参数 fetches 接收的 accuracy 不同的名字即可，如下：

accuracy_result = sess.run(accuracy, feed_dict={images_holder: batch_images.eval(), labels_holder: batch_labels}