Skip to content

Merging the latent diffusion model#2278

Draft
Jubeku wants to merge 398 commits into
developfrom
jk/develop/diffusion-full-pipeline
Draft

Merging the latent diffusion model#2278
Jubeku wants to merge 398 commits into
developfrom
jk/develop/diffusion-full-pipeline

Conversation

@Jubeku
Copy link
Copy Markdown
Contributor

@Jubeku Jubeku commented Apr 28, 2026

Description

Draft PR to merge the latent diffusion model into develop.

TODO:

  • Update config for diffusion parameters similar to SSL parameters.

Issue Number

Closes #2277

Is this PR a draft? Mark it as draft.

Checklist before asking for review

  • I have performed a self-review of my code
  • My changes comply with basic sanity checks:
    • I have fixed formatting issues with ./scripts/actions.sh lint
    • I have run unit tests with ./scripts/actions.sh unit-test
    • I have documented my code and I have updated the docstrings.
    • I have added unit tests, if relevant
  • I have tried my changes with data and code:
    • I have run the integration tests with ./scripts/actions.sh integration-test
    • (bigger changes) I have run a full training and I have written in the comment the run_id(s): launch-slurm.py --time 60
    • (bigger changes and experiments) I have shared a hegdedoc in the github issue with all the configurations and runs for this experiments
  • I have informed and aligned with people impacted by my change:
    • for config changes: the MatterMost channels and/or a design doc
    • for changes of dependencies: the MatterMost software development channel

shmh40 and others added 30 commits November 28, 2025 11:34
Had placeholder for testing, now back to "random" for masking strategy in the base level of default_config
* noise generation in diffusion inference

* lint

---------

Co-authored-by: Matthias Karlbauer <matthias.karlbauer@ecmwf.int>
…eatherGenerator into shmh40/dev/1270-idx-global-local
* Add to device to ModelBatch, etc & adapt model

TODO adapt validate and inference
TODO test forecasting and multiple stream because predict changed
substantially

* Rename view to sample and fix validate

* Revert predict function and fix inference

* Fix invalid access with mask

* Linting

* Fixed handling of target_idxs and other minor issues

---------

Co-authored-by: sophiex <24638638+sophie-xhonneux@users.noreply.github.com>
Co-authored-by: Christian Lessig <christian.lessig@ecmwf.int>
moritzhauschulz and others added 30 commits May 18, 2026 19:18
…peline-forecast' into jk/develop/diffusion-full-pipeline
* additional check for num_input_steps

* changed configs and fixed error in inference_forward

* review PR comments

* comma

* uncommented roll-out conditioning

* update parameters to freeze by default

* included l3rxe29i as default loaded model

* rm breakpoint

* review PR

* config changes
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

eval anything related to the model evaluation pipeline infra Issues related to infrastructure model Related to model training or definition (not generic infra)

Projects

Status: No status

Development

Successfully merging this pull request may close these issues.

Diffusion model merging draft PR

10 participants