python - 类型错误 : Cannot interpret feed_dict key as Tensor: Can not convert a int into a Tensor

Question

以下是我使用tensorflow的代码，我已经尝试调试这个错误两天了......

changed_result = y_conv.eval(feed_dict={x: pixelNew, keep_prob: 1.0})

和:

changed_result = tf.Variable([0.0],tf.float32)

我尝试将changed_result设为张量，或者只是不声明它。两者都不起作用。

以下是完整的错误描述:

Traceback (most recent call last): File "deepMnist.py", line 235, in tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)

File "/Users/ximinlin/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/platform/app.py", line 43, in run sys.exit(main(sys.argv[:1] + flags_passthrough))

File "deepMnist.py", line 162, in main changed_result = y_conv.eval(feed_dict={x: pixelNew, keep_prob: 1.0}) File "/Users/ximinlin/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 575, in eval return _eval_using_default_session(self, feed_dict, self.graph, session)

File "/Users/ximinlin/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 3633, in _eval_using_default_session return session.run(tensors, feed_dict)

File "/Users/ximinlin/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 766, in run run_metadata_ptr)

File "/Users/ximinlin/anaconda/envs/tensorflow/lib/python2.7/site-packages/tensorflow/python/client/session.py", line 921, in _run + e.args[0]) TypeError: Cannot interpret feed_dict key as Tensor: Can not convert a int into a Tensor.

以下是x和keep_pro的定义:

x = tf.placeholder(tf.float32, [None, 784])

keep_prob = 0

keep_prob和x实际上是tensorflow教程中的代码，我只是将它们复制到我的代码中

Answer 1

您没有显示足够的上下文来给出非常具体的答案，但您的错误是 x 或 keep_prob 不是张量，而是一个整数，其中你调用它(如果你之前将它们定义为张量，你可能已经覆盖它们)。

因此，错误在于 x 或 keep_prob，而不是 changed_result。