Repository navigation
Stop inverse-square-root cooldowns at zero - #220
Merged
jpuigcerver merged 1 commit intoOct 15, 2026
Merged
jpuigcerver merged 1 commit into
jpuigcerver merged 1 commit into
Conversation
sylvesterkaczmarek
marked this pull request as ready for review
September 22, 2026 22:33
sylvesterkaczmarek
force-pushed
the
fix/stop-rsqrt-cooldown-at-zero
branch
from
September 30, 2026 14:17
d894ea7 to
eab2dab
Compare
Contributor
Author
|
Could a maintainer review the learning-rate cooldown endpoint fix and unblock the workflow approval and Copybara import? |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Hold the inverse-square-root schedule at zero at and after the end of an enabled cooldown. The current linear expression continues below zero: with
peak_value=0.1,decay_steps=12,warmup_steps=4,cooldown_steps=3, andtimescale=2, step 13 produces approximately-0.0178174. An optimizer that continues past this boundary resumes parameter updates with a reversed learning-rate sign.Use an explicit endpoint selection so the result is exactly zero in eager and JIT execution. Preserve the existing warmup, inverse-square-root phase, and in-range cooldown expression. Schedules without a cooldown continue their inverse-square-root tail as before.
Validation
Python 3.12.11, JAX 0.11.2, Flax 0.12.9, Optax 0.2.8, NumPy 2.5.3, macOS arm64 CPU:
6dcc8c0: ten new regression cases fail and two controls pass. All 12 pass with the fix.JAX_ENABLE_X64=1.git diff --checkpass.The full data/checkpoint/training suite, end-to-end image training, and accelerator/multihost execution were not run. This branch changes different files and is independent of the pending weight-decay PR #219.