Skip to content

Default max-context-length of 4096 for LLM iOS platform exports#81

Merged
Lewis300 merged 2 commits into
apple:mainfrom
Lewis300:fix-ios-max-ctx-len
Jul 2, 2026
Merged

Default max-context-length of 4096 for LLM iOS platform exports#81
Lewis300 merged 2 commits into
apple:mainfrom
Lewis300:fix-ios-max-ctx-len

Conversation

@Lewis300

@Lewis300 Lewis300 commented Jul 2, 2026

Copy link
Copy Markdown
Contributor
  • Default max-context-length of 4096 for LLM iOS platform exports including experimental models not included in the model registry
  • Add check to ensure that the max context length in the ExportConfig is less than or equal to that of the HF config
  • Add tests to ensure this new infra functions as expected

@Lewis300 Lewis300 requested review from DawerG, stikves and tjia1818 July 2, 2026 21:26
Comment thread python/src/coreai_models/export/pipeline.py
@Lewis300 Lewis300 force-pushed the fix-ios-max-ctx-len branch 6 times, most recently from b227563 to 660dd96 Compare July 2, 2026 22:08
… user-specified max-context-lengths are <= to the HF max_position_embeddings
@Lewis300 Lewis300 force-pushed the fix-ios-max-ctx-len branch from 660dd96 to 44c331b Compare July 2, 2026 22:09
@Lewis300 Lewis300 merged commit e203a0d into apple:main Jul 2, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants