TypeError: Descriptors cannot not be created directly.（Tensorflow安装问题）

miko-N＿N-

于 2023-01-15 13:04:54 发布

阅读量3.7k

点赞数

文章标签： tensorflow python 深度学习

本文链接：https://blog.youkuaiyun.com/qq_45738555/article/details/128693591

版权

在调试Tacotron2模型时遇到tensorflow导入错误，问题源于protobuf版本过高。通过降级protobuf到3.20.x解决了此问题，之前还曾因tensorflow与CUDA版本不兼容将CUDA从12降至10.2。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

笔者在调试Tacotron2官模时，调取tensorflow时出现了这样的错误。

Traceback (most recent call last):
  File "M:/project/project/TTS/Offical_model/Tacotron2/tacotron2/junk/test/torch_version_test.py", line 8, in <module>
    import tensorflow
  File "M:\project\env\anaconda\env\Tacotron2\lib\site-packages\tensorflow\__init__.py", line 99, in <module>
    from tensorflow_core import *
  File "M:\project\env\anaconda\env\Tacotron2\lib\site-packages\tensorflow_core\__init__.py", line 28, in <module>
    from tensorflow.python import pywrap_tensorflow  # pylint: disable=unused-import
  File "<frozen importlib._bootstrap>", line 1019, in _handle_fromlist
  File "M:\project\env\anaconda\env\Tacotron2\lib\site-packages\tensorflow\__init__.py", line 50, in __getattr__
    module = self._load

最低0.47元/天解锁文章