Change the default optimizer from signum to adam - #139
Merged
Conversation
Adam is the recipe our released SAEs were actually trained with (e.g. sae-SmolLM2-1.7B-layer17-32x used adam at lr 1e-4), so the default should match what we publish rather than requiring an explicit flag. The per-optimizer auto-LR in Trainer already handles the switch: adam picks 2e-4 / sqrt(num_latents / 2**14), which for a 32x SAE on d_model=2048 (65,536 latents) is exactly the 1e-4 used for those runs. Adam also has an lr warmup schedule wired up, which signum (ScheduleFree) does not. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This is the Schelling point and I disagree with our earlier decision to use something else