Skip to content
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

Rework ID-based parametrisation #27

Merged
merged 4 commits into from
Sep 11, 2024
Merged

Rework ID-based parametrisation #27

merged 4 commits into from
Sep 11, 2024

Conversation

wesselb
Copy link
Contributor

@wesselb wesselb commented Sep 9, 2024

This should make it easier to extend the model to new variables.

  • Check correctness in notebooks. Looks good.
  • Remove warnings.
  • Add example in docs.

Closes #24

@wesselb wesselb self-assigned this Sep 9, 2024
docs/finetuning.md Outdated Show resolved Hide resolved
Copy link
Contributor

@a-lucic a-lucic left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Happy this is implemented!

@scottcha
Copy link
Contributor

I checked this PR against my fine tuning and was able to add variables w/out needing to directly modify the layers (all completed on the Small version of Aurora). Adding the ERA5 tsn var w/out any additional fine tuning results in
image

After a few mins of fine tuning I get this for tsn
image

For the existing 2t var I get this w/out finetuning
image

And this after a few mins of fine tuning
image

While I'm not super experienced yet with leveraging this model so far the behavior seems to be expected and its possible to add vars more easily than previously.

@wesselb
Copy link
Contributor Author

wesselb commented Sep 11, 2024

Thanks, @scottcha! I'm going to interpret this as evidence that the refactoring works as intended and proceed to merging. :)

@wesselb wesselb merged commit b65b87d into main Sep 11, 2024
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Adding a var causes incorrect output
3 participants