报错Failed precondition: Attempting to use uninitialized value wavelet/n_l

Question

我初始化了变量

n_h = tf.get_variable('n_h', initializer=0.0, trainable=trainable_set)

但是总是在训练的时候报错：

Traceback (most recent call last):
  File "D:\video_watermarking\code\ReDMark\Train_harr3_lifting_GT.py", line 171, in <module>
    encoder_input = layers.Concatenate(axis=-1, name='encoder_input')([input_dwt_img, input_watermark])  # 4,4,64
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer.py", line 802, in __call__
    base_layer_utils.create_keras_history(inputs)
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer_utils.py", line 184, in create_keras_history
    _, created_layers = _create_keras_history_helper(tensors, set(), [])
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer_utils.py", line 231, in _create_keras_history_helper
    layer_inputs, processed_ops, created_layers)
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer_utils.py", line 231, in _create_keras_history_helper
    layer_inputs, processed_ops, created_layers)
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer_utils.py", line 231, in _create_keras_history_helper
    layer_inputs, processed_ops, created_layers)
  [Previous line repeated 5 more times]
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\engine\base_layer_utils.py", line 229, in _create_keras_history_helper
    constants[i] = backend.function([], op_input)([])
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\keras\backend.py", line 3476, in __call__
    run_metadata=self.run_metadata)
  File "D:\program\anaconda3\envs\tensorflow1_15_gpu\lib\site-packages\tensorflow_core\python\client\session.py", line 1472, in __call__
    run_metadata_ptr)
tensorflow.python.framework.errors_impl.FailedPreconditionError: 2 root error(s) found.
  (0) Failed precondition: Attempting to use uninitialized value wavelet/n_l
     [[{{node wavelet/n_l/read}}]]
     [[wavelet/add_79/_25]]
  (1) Failed precondition: Attempting to use uninitialized value wavelet/n_l
     [[{{node wavelet/n_l/read}}]]
0 successful operations.
0 derived errors ignored.

这是为什么呢？

Answer 1

【以下回答由 GPT 生成】

问题解决方案:

根据报错信息，可以看到报错是由于使用了tf.get_variable函数导致的。在TensorFlow 2.0版本中，该函数已经不再被推荐使用，因此可以尝试使用tf.Variable函数替代。

以下是修改后的代码片段：

n_h = tf.Variable(initial_value=0.0, trainable=trainable_set, name='n_h')

修改后的代码使用了tf.Variable函数来初始化变量n_h，并为其添加了一个name参数。

希望以上解决方案能够帮助到您，如果有任何问题，请随时提问。

如果你已经解决了该问题, 非常希望你能够分享一下解决方案, 写成博客, 将相关链接放在评论区, 以帮助更多的人 ^-^