Summary
Add async embedding support for the Twelve Labs provider to handle audio/video content up to 4 hours.
Context
Phase 16 implements the Twelve Labs embedding function using the sync endpoint (POST /v1.3/embed-v2), which handles text, image, and audio/video under 10 minutes. The async endpoint (POST /v1.3/embed-v2/tasks) supports audio/video up to 4 hours but requires task creation and polling for completion.
Requirements
- Add async task creation via
POST /v1.3/embed-v2/tasks
- Implement polling loop to check task status and retrieve results
- Support audio/video content up to 4 hours
- Integrate with the existing
ContentEmbeddingFunction interface (or extend if needed)
References
Summary
Add async embedding support for the Twelve Labs provider to handle audio/video content up to 4 hours.
Context
Phase 16 implements the Twelve Labs embedding function using the sync endpoint (
POST /v1.3/embed-v2), which handles text, image, and audio/video under 10 minutes. The async endpoint (POST /v1.3/embed-v2/tasks) supports audio/video up to 4 hours but requires task creation and polling for completion.Requirements
POST /v1.3/embed-v2/tasksContentEmbeddingFunctioninterface (or extend if needed)References