Add OLMo 2 support (1B, macOS + iOS) - #232
Merged
Merged
Conversation
stikves
force-pushed
the
sukru/olmo2-support
branch
from
September 10, 2026 00:41
546c24c to
de1b526
Compare
stikves
marked this pull request as ready for review
September 10, 2026 00:44
stikves
requested review from
DawerG,
Lewis300,
gokulkrishna98,
kevchengcodes,
pkmandke and
tjia1818
September 10, 2026 00:44
tjia1818
approved these changes
Sep 10, 2026
pkmandke
reviewed
Sep 10, 2026
pkmandke
reviewed
Sep 10, 2026
Co-authored-by: Prathamesh <46148373+pkmandke@users.noreply.github.com>
group_size 8 was unnecessarily fine; a wikitext PPL sweep at the model's native 4096 context shows 8/16/32 are equivalent within noise (17.18 / 17.09 / 17.23). group_size 16 halves the palettization LUT overhead versus 8 while holding quality (17.09). Update the README iOS 6-bit perplexity accordingly.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
x + norm(attn(x))) and flat QK-norm (RMSNorm over all head dims jointly before reshape)Evaluation
word_perplexityon WikiText-2, 4K context.float16)float16)Test plan