Re-cluster palettization centroids from the warmed-up weights when a PATSchedule enables fake palettization mid-training - #86
Merged
crowbat merged 3 commits intoSep 4, 2026
Conversation
anotheranshu
reviewed
Sep 3, 2026
| new_state = schedule._compute_state(self._step_count) | ||
| # On the warm-up -> enabled transition, re-cluster from the current | ||
| # (warmed-up) weights instead of the frozen prepare-time centroids. | ||
| if new_state and not fp_module.fake_palett_enabled[0]: |
Contributor
There was a problem hiding this comment.
Have you considered pushing the call to mark the centroids stale down into enable_fake_palett? That way it's not tied exclusively to the scheduler; if the user directly enables the fake palettizer we can trigger a centroid update.
…PATSchedule enables fake palettization mid-training
crowbat
force-pushed
the
u/k_hsieh/recompute_centroids_when_enabling_fake_palettize_for_pat_schedule
branch
from
September 3, 2026 21:05
83b3b6b to
5557765
Compare
anotheranshu
approved these changes
Sep 3, 2026
crowbat
deleted the
u/k_hsieh/recompute_centroids_when_enabling_fake_palettize_for_pat_schedule
branch
September 4, 2026 00:47
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Currently if enable_fake_palettize > 0 in PATSchedule, centroids are initialized at the start and reused when the module is later enabled via PATSchedule and palettizer.step(). Instead, recompute centroids based on trained updated weights that changed during step < enable_fake_palettize.