Since I haven’t installed python manually, why are there so many versions installed?
Can I clean it all up and keep only the latest one?
How do you uninstall older versions?
What are the consequences?
I do this request because when I retrieve third-party tools or make updates, I never know which version of pip Ihave to use. I’m also having problems with version conflicts and dependencies (especially with the latest version of openIA whisper).
Those are all the same, more correctly: they point to the same version. You can check e.g. with python --version.
They are there so that you can address a specific version (in your scripts) if you need to, but can use convenient shortcuts when you want or if yoy do not care about the specific version in your script.
I’ve been wondering about the number of different versions of python for a while, and I didn’t really understand how it worked. I understand better now.
Concerning Whisper, I’m having problems with conflicts when updating or reinstalling. But these don’t seem to come from python. OpenIA provides a link but I don’t understand (English is not my native tongue, and Idont understand explanations).
The conflict is caused by:
openai-whisper 20231106 depends on triton==2.0.0
openai-whisper 20231105 depends on triton==2.0.0
openai-whisper 20230918 depends on triton==2.0.0
openai-whisper 20230314 depends on triton==2.0.0
openai-whisper 20230308 depends on triton==2.0.0
openai-whisper 20230307 depends on triton==2.0.0
openai-whisper 20230306 depends on triton>=2.0.0.dev20221202
openai-whisper 20230124 depends on torch
openai-whisper 20230117 depends on torch
To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict
ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts
'm not sure that using a virtual environment will solve the problem (and I wouldn’t know how to do it anyway).
If this succeeds, then there’s likely something wrong with the Python packages that you installed locally / globally for your account (which is why virtualenvs are a good idea).
Python and venv are very much a fedora related topic. fedora 39 contained python 3.11 while fedora 40 upgraded to python 3.12. The changes and the use of a venv to use different versions of python are very much related to fedora directly.