Closed melioristic closed 2 years ago
Hi Mohit, You've created a conda environment and installed packages using pip!?
To export the requirements.txt file you use (with the environment activated)
pip freeze > requirements.txt
From past experience and especially with ML packages like pytorch and keras, I would recommend that stick with just conda packages from conda-forge. So , I would recreate the environment and install packages with (again with the environment activated)
conda install -c conda-forge numpy pytorch (...)
Then to export the environment to a file you use
conda env export > environment.yaml
A useful page is
Hope this helps
Juan
Em sáb., 14 de mai. de 2022 18:47, Mohit Anand @.***> escreveu:
I created the conda environment and installed the packages using the pip command. Is there some way to create the environment file that includes the packages installed through the pip command? We need to look at some of the most reproducible ways to create the environment. @colonesej https://github.com/colonesej Do you have some suggestions on this?
— Reply to this email directly, view it on GitHub https://github.com/esowc/CW4Floods/issues/1#issuecomment-1126783864, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALK7JOGAEVLQQENKEJ7VQOTVJ7RLHANCNFSM5V3VYP2A . You are receiving this because you were mentioned.Message ID: @.***>
So I did create the new environment added everything through conda-forge except xarray, which was installed using conda install -c anaconda xarray
Fixed through the commit. https://github.com/esowc/CW4Floods/commit/8cb0df6ed258f28bfdcb9af8967aef899be28263
I created the conda environment and installed the packages using the pip command. Is there some way to create the environment file that includes the packages installed through the pip command? We need to look at some of the most reproducible ways to create the environment. @colonesej Do you have some suggestions on this?