Roy Belio

### Is there an existing issue for this? - [X] Yes, I have searched the existing issues - [X] Yes, I have checked troubleshooting page ### The problem Using windows...

Windows 10/11 are tracking app data usage for each network device. this can be disabled by change the key `Computer\HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Services\Ndu`:`Start` to the value 4(DWORD) Further information can be found here:...

Remove scipy dependency

4

Its a heavy dependency which requires external libraries and tooling installed. We are only using it for the scipy.ndimage.zoom method which is provided both by Pillow and OpenCV which are...

fix(nvidia-safety): correct NeMo Guardrails API endpoint

5

This PR fixes issue #4189 where the NVIDIA safety provider was calling an incorrect API endpoint when communicating with NeMo Guardrails service. ## Problem The NVIDIA safety provider implementation was...

Fix AttributeError in streaming response cleanup

7

This PR fixes issue #3185 The code calls `await event_gen.aclose()` but OpenAI's `AsyncStream` doesn't have an `aclose()` method - it has `close()` (which is async). when clients cancel streaming requests,...

feat(cli): use gunicorn to manage server workers on unix systems

14

# What does this PR do? This PR adds production-grade server capabilities to Llama Stack by integrating Gunicorn with Uvicorn workers on Unix-based systems (Linux, macOS). The implementation provides multi-process...

# What does this PR do? This PR fixes an issue where custom collection names for vector stores were not being correctly utilized by the underlying storage providers, resulting in...