Dockerfile should use the correct nvidia channel to install libcu* #423
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: Praateek <praateekm@gmail.com>
sarahyurick
added
gpuci
sarahyurick
requested changes
Dec 11, 2024
| # Install the minimal libcu* libraries needed by NeMo Curator | ||
| RUN conda create -y --name curator -c conda-forge -c nvidia \ | ||
| RUN conda create -y --name curator -c nvidia/label/cuda-${CUDA_VER} -c conda-forge | ||
| python=3.10 \ |
Contributor
There was a problem hiding this comment.
Looks like gpuCI is currently throwing an error here.
Contributor
Author
There was a problem hiding this comment.
Apparently build arg's can't be in RUN, so had to create env var, which is ugly
Contributor
There was a problem hiding this comment.
Oh lol, interesting. LGTM!
praateekmahajan
added
gpuci
Signed-off-by: Praateek <praateekm@gmail.com>
praateekmahajan
added
gpuci
sarahyurick
approved these changes
Dec 11, 2024
| # Install the minimal libcu* libraries needed by NeMo Curator | ||
| RUN conda create -y --name curator -c conda-forge -c nvidia \ | ||
| RUN conda create -y --name curator -c nvidia/label/cuda-${CUDA_VER} -c conda-forge | ||
| python=3.10 \ |
Contributor
There was a problem hiding this comment.
Oh lol, interesting. LGTM!
jnke2016
pushed a commit
to jnke2016/Curator
that referenced
this pull request
Nov 12, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Realised that while we're installing cuda-12.5.1 in the CI image, the conda installations are all over the place and some of them are being picked up from conda-forge rather than nvidia. (IIUC -c A -c B, A takes precedence over B, and in our case conda-forge might have older versions of packages)
Secondly if we wish to use UDFs on
dask_cudfwe needNVVM. The place where I needed it was hack for #417Checklist