class ArgumentParser(Generic[_T], metaclass=_ArgumentParserCache): TypeError: metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases

Question

class ArgumentParser(Generic[_T], metaclass=_ArgumentParserCache): TypeError: metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases

Closed this issue 2 months ago · 3 comments

I install tensorflow=1.14 along with cudatoolkit=10.0 in a new conda environment. on installing and restarting the system I tried to import tensorflow but I got err from:
abls\flags_argument_parser.py
class ArgumentParser(Generic[_T], metaclass=_ArgumentParserCache):
TypeError: metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases

I checked th version of absl-py it is 2.0.0
What can be done?

Answer 1 · 2024-04-02T18:23:17.000Z

Which Python version are you using? It can be an error in Python 3.6 or earlier, and absl-py 2.0.0 has dropped support for Python 3.6.

Answer 2 · 2024-04-02T18:27:54.000Z

Yes I have installed python version 3.6 in the environment.

Answer 3 · 2024-04-02T18:29:50.000Z

You'll either need to use newer Python versions or downgrade absl-py to pre 2.0.0.