AttributeError: 'Tensor' 对象没有属性'_in_graph_mode'。

Posted 2021-03-19

我遇到一个错误：'Tensor'对象没有属性'_in_graph_mode'。我调试了一下代码，我认为是在这个 渐变带 功能，但我不知道为什么。如果有谁知道，请帮帮我! :)

for i in range(50):
     with tf.GradientTape() as tape:
          inverted_feature = tf.cast(opt_img, dtype)
          content_feature = tf.cast(images, dtype)
          conv_inverted_outputs = grad_model(inverted_feature)
          conv_content_outputs = grad_model(content_feature)
          loss = InvertedImage.get_loss(conv_content_outputs, conv_inverted_outputs, content_feature, inverted_feature)

哪儿 毕业生模型 取特定层的输入和输出。此外。opt_img 和*图像*是张力器

     grads = tape.gradient(loss, [conv_inverted_outputs, conv_content_outputs])

     processed_grads = [g for g in grads]

     grads_and_vars = zip(processed_grads, [conv_inverted_outputs, conv_content_outputs])
     opt.apply_gradients(grads_and_vars)

我得到这个错误。

Traceback (most recent call last):
  File "/usr/local/lib/python3.7/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/local/lib/python3.7/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/helena/.vscode/extensions/ms-python.python-2020.2.64397/pythonFiles/lib/python/new_ptvsd/wheels/ptvsd/__main__.py", line 45, in <module>
    cli.main()
  File "/home/helena/.vscode/extensions/ms-python.python-2020.2.64397/pythonFiles/lib/python/new_ptvsd/wheels/ptvsd/../ptvsd/server/cli.py", line 361, in main
    run()
  File "/home/helena/.vscode/extensions/ms-python.python-2020.2.64397/pythonFiles/lib/python/new_ptvsd/wheels/ptvsd/../ptvsd/server/cli.py", line 203, in run_file
    runpy.run_path(options.target, run_name="__main__")
  File "/usr/local/lib/python3.7/runpy.py", line 263, in run_path
    pkg_name=pkg_name, script_name=fname)
  File "/usr/local/lib/python3.7/runpy.py", line 96, in _run_module_code
    mod_name, mod_spec, pkg_name, script_name)
  File "/usr/local/lib/python3.7/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/helena/Documents/LAR_Celesc/lar-computer-vision/objdet-api/test_inverted_image.py", line 20, in <module>
    data, model, class_index=tabby_cat_class_index, layer_name="block5_conv3"
  File "/home/helena/Documents/LAR_Celesc/lar-computer-vision/objdet-api/tf_explain/core/inverted_image.py", line 54, in explain
    images, model, class_index, layer_name
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/def_function.py", line 568, in __call__
    result = self._call(*args, **kwds)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/def_function.py", line 615, in _call
    self._initialize(args, kwds, add_initializers_to=initializers)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/def_function.py", line 497, in _initialize
    *args, **kwds))
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/function.py", line 2389, in _get_concrete_function_internal_garbage_collected
    graph_function, _, _ = self._maybe_define_function(args, kwargs)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/function.py", line 2703, in _maybe_define_function
    graph_function = self._create_graph_function(args, kwargs)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/function.py", line 2593, in _create_graph_function
    capture_by_value=self._capture_by_value),
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/framework/func_graph.py", line 978, in func_graph_from_py_func
    func_outputs = python_func(*func_args, **func_kwargs)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/eager/def_function.py", line 439, in wrapped_fn
    return weak_wrapped_fn().__wrapped__(*args, **kwds)
  File "/home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/framework/func_graph.py", line 968, in wrapper
    raise e.ag_error_metadata.to_exception(e)
AttributeError: in converted code:

    /home/helena/Documents/LAR_Celesc/lar-computer-vision/objdet-api/tf_explain/core/inverted_image.py:125 get_optimize_image  *
        opt.apply_gradients(grads_and_vars)
    /home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/keras/optimizer_v2/optimizer_v2.py:434 apply_gradients
        self._create_slots(var_list)
    /home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/keras/optimizer_v2/gradient_descent.py:100 _create_slots
        self.add_slot(var, "momentum")
    /home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/keras/optimizer_v2/optimizer_v2.py:574 add_slot
        var_key = _var_key(var)
    /home/helena/Documents/LAR_Celesc/larenv/lib/python3.7/site-packages/tensorflow_core/python/keras/optimizer_v2/optimizer_v2.py:1065 _var_key
        if var._in_graph_mode:

    AttributeError: 'Tensor' object has no attribute '_in_graph_mode'

'''

答案

我刚刚遇到类似的问题（同样的跟踪），并找到了一个修复方法，希望对你的情况也有帮助。

检查一下

conv_inverted_outputs, conv_content_outputs

类型是tf. Variable而不是tf. Tensor.

如果其中任何一个是tf.Tensor，你会得到这个错误。

我没有看到你的代码，无法给出确切的修复方法，但在我的案例中，我是用以下方式生成权重的。

weights = tf.random.normal((784, 10))

改成:

weights = tf.Variable(tf.random.normal((784, 10)))

就解决了这个问题