Skip to content

Remove filtering of q_cells and identity in ema calculation#2602

Draft
shmh40 wants to merge 1 commit into
developfrom
shmh40/dev/reenable-q-cells-ema
Draft

Remove filtering of q_cells and identity in ema calculation#2602
shmh40 wants to merge 1 commit into
developfrom
shmh40/dev/reenable-q-cells-ema

Conversation

@shmh40

@shmh40 shmh40 commented Jul 6, 2026

Copy link
Copy Markdown
Contributor

Removed condition to skip parameters containing 'identity' or 'q_cells' in ema calculation.

Only tested for JEPA without FSDP, not tested in any other contexts.

Description

See above.

Issue Number

Closes #2601

Is this PR a draft? Mark it as draft.

Checklist before asking for review

  • I have performed a self-review of my code
  • [ x 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

Removed condition to skip parameters containing 'identity' or 'q_cells' in ema calculation
@shmh40 shmh40 self-assigned this Jul 6, 2026
@shmh40 shmh40 added the model Related to model training or definition (not generic infra) label Jul 6, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

model Related to model training or definition (not generic infra)

Projects

Status: No status

Development

Successfully merging this pull request may close these issues.

Re-enable q_cells being included in ema computation

1 participant