ISSUE 534: Deprecate conda environment.yml / move to pip#552
ISSUE 534: Deprecate conda environment.yml / move to pip#552HundredBillion wants to merge 12 commits intosdatkinson:mainfrom
Conversation
|
Thanks for the PR, @HundredBillion! This doesn't resolve #534 because installing still uses conda .yml files. I'll give comments inline presently |
|
@sdatkinson - I noticed your PR to change the url for installing Pytorch. I updated this PR in case you still want to use update the docs to install via |
sdatkinson
left a comment
There was a problem hiding this comment.
I'm sorry, but this still isn't what I want. There are yml files still in the repo. Moving to pip should see them gone.
Got it — I misunderstood the issue. I treated it as “use pip inside the conda env files,” but what you want is to remove the conda env approach entirely. I’ll update the PR to delete the environment .yml files and switch the installation docs to a pip/venv-based flow. |
There was a problem hiding this comment.
Pull request overview
This PR updates the project’s installation guidance to move away from conda-based environment setup and toward pip/venv-based installation, addressing Issue #534 about PyTorch no longer being available via conda.
Changes:
- Removed conda environment definitions for CPU (Apple) and GPU setups.
- Rewrote the installation docs to use Python +
venv+python -m pip. - Updated local development setup instructions to use editable installs and manual dev tooling installation.
Reviewed changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated 4 comments.
| File | Description |
|---|---|
| environments/environment_gpu.yml | Removes the GPU conda environment definition. |
| environments/environment_cpu_apple.yml | Removes the Apple/CPU conda environment definition. |
| docs/source/installation.rst | Updates installation and development instructions to use venv + pip and points users to PyTorch’s install selector. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
This PR is meant to resolve Issue 534.
Since PyTorch is no longer shipping conda packages, I updated this repo to remove the conda environment
.ymlfiles and switch the installation docs over to apip/ virtualenv-based setup.For GPU installs, PyTorch’s commands depend on the user’s OS and CUDA version, so instead of hardcoding those here, the docs now point to the official PyTorch install page:
https://pytorch.org/get-started/locally/
After PyTorch is installed, NAM can be installed with
pip.Summary of changes
environments/environment_cpu_apple.ymlenvironments/environment_gpu.ymlrequirements-gpu.txtdocs/source/installation.rstto describe a pip/venv-based install flow