-
Notifications
You must be signed in to change notification settings - Fork 14
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
pip install has issues detecting dependencies on conda #6
Comments
This is a known issue with poetry A possible workaround can be found here: cupy/cupy#4727 (comment)
|
Not sure why pip install --no-build-isolation have following issue:
But when I install 1.0.4 no issue above. |
Hi @Kin-Zhang, It seems like you are installing FastGeodis from cache, which means it is using a previously built version from the cache that possibly was built with another setup on your machine. Could you try to force a rebuild:
|
If a system has pytorch pre-installed using conda - then poetry (pyproject.toml) does not detect it and installs pytorch using pip
This causes issues especially if there is a mismatch between CUDA version for pytorch from pip and the CUDA version available on system
To reproduce:
docker run --gpus all -it --rm nvcr.io/nvidia/pytorch:21.10-py3
pip install FastGeodis
The above gives the following error:
The text was updated successfully, but these errors were encountered: