Check torch and lightning versions before running models #444
Conversation
This file contains 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
Borda
reviewed
Aug 22, 2024
Borda
approved these changes
Aug 22, 2024
Co-authored-by: Jirka Borovec <[email protected]>
Co-authored-by: Jirka Borovec <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
I think that a lot of weird errors with uninformative error messages that users encounter (e.g., see #438) are due to package mismatches. Since we are using the latest features from torch and Fabric, code kind of runs but then there's this weird error here and there, or even a missing feature.
So, I propose only allowing to run the scripts if the minimal versions are installed. If a user doesn't have these required PyTorch and Lightning versions, there is a message informing the user about that.