Skip to content

[pull] main from inclusionAI:main#26

Merged
pull[bot] merged 1 commit intoaxistore80-coder:mainfrom
inclusionAI:main
Apr 2, 2026
Merged

[pull] main from inclusionAI:main#26
pull[bot] merged 1 commit intoaxistore80-coder:mainfrom
inclusionAI:main

Conversation

@pull
Copy link
Copy Markdown

@pull pull bot commented Apr 2, 2026

See Commits and Changes for more details.


Created by pull[bot] (v2.0.0-alpha.4)

Can you help keep this open source service alive? 💖 Please sponsor : )

* refactor(archon): unify inference service workflows and add online rollout support

Replace OpenAIProxyWorkflow with InferenceServiceWorkflow that supports both offline (agent-driven) and online (HITL/callback-driven) rollout modes. Remove explicit end_session endpoint in favor of automatic session finalization via set_reward.

Key changes:

- Add InferenceServiceWorkflow with offline and online modes

- Remove /rl/end_session endpoint; finalize sessions via set_reward

- Move admin_api_key from GatewayControllerConfig into OpenAIProxyConfig

- Add callback server for online trajectory ready notifications

- Unify SessionData to support both batch and HITL sessions

- Add sticky HITL routing in router (replaces admin round-robin)

- Add online rollout example scripts and integration tests

- Use IPv6-safe address formatting throughout guard and data proxy

* revert: drop openai proxy test changes from this PR

* docs(examples): expand inference service example guide

Document both the offline rollout and HITL online demo flows so users can run the new inference-service examples without reading the implementation first.

Key changes:
- describe the offline tau2 rollout example
- add setup and usage notes for the HITL online demo
- document the manual reward submission flow

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* feat(examples): add HITL demo with Qwen3-0.6B default

Add the Python human-in-the-loop inference-service demo and document how to run it with the public Qwen/Qwen3-0.6B model identifier.

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* refactor(examples): remove legacy online demo helper scripts

Drop the retired shell-based online demo utilities now that the Python human-in-the-loop demo covers the supported workflow and no longer needs the old helper modules.

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
@pull pull bot locked and limited conversation to collaborators Apr 2, 2026
@pull pull bot added the ⤵️ pull label Apr 2, 2026
@pull pull bot merged commit 9332183 into axistore80-coder:main Apr 2, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant