Add chat template kwargs to chat completions - #1009
Draft
jennyf19 wants to merge 5 commits into
Draft
Conversation
|
Someone is attempting to deploy a commit to the MSFT-AIP Team on Vercel. A member of the Team first needs to authorize it. |
added 5 commits
August 18, 2026 09:05
Preserve typed JSON values from the OpenAI-compatible request through text, media, and continued-turn prompt rendering. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a9f5bae1-1b88-4311-8fac-6224769e0400
Build the dependency stack only inside Jenny's fork and run the affected Foundry C++ unit suites without changing production dependency pins. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a9f5bae1-1b88-4311-8fac-6224769e0400
Foundry's manifest pins an older baseline, so the disposable workflow must retain version-history commits and install Ninja explicitly. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a9f5bae1-1b88-4311-8fac-6224769e0400
The dependency stack compiled and affected Foundry unit suites passed in fork run 32079437566. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a9f5bae1-1b88-4311-8fac-6224769e0400
Run the stacked native build and affected Foundry tests after rebasing onto the shared preprocessing architecture. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: a9f5bae1-1b88-4311-8fac-6224769e0400
jennyf19
force-pushed
the
feature/chat-template-kwargs
branch
from
August 18, 2026 16:06
883bebe to
1a9ac05
Compare
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.
Summary
chat_template_kwargsto the OpenAI-compatible chat completions requestDependencies
This draft depends on:
No personal-fork dependency pins are present in this PR. Keep this PR draft until both upstream APIs merge and Foundry updates its normal ORT GenAI package version.
Motivation
Qwen 3's bundled chat template defaults to reasoning. Without
enable_thinking:false, bounded requests can consume their entire completion budget inside<think>and never produce an answer or tool call. Qwen 2.5 does not have this reasoning-template branch.This enables clients to send:
{ "chat_template_kwargs": { "enable_thinking": false } }Related: #808
Validation
A disposable fork workflow built the complete stack—ORT Extensions #1102, ORT GenAI #2421, and this branch—without changing production dependency pins. The affected Foundry C++ unit suites passed:
https://github.com/jennyf19/foundry-local/actions/runs/32079437566