llama.cpp UI adds HuggingFace Hub model discovery and downloads

The llama.cpp UI now has a Discover Models view that searches and filters GGUF models on HuggingFace Hub. Downloading a quantized model automatically queues its draft sidecar files (dflash, mtp), with pause, resume and cancel applied synchronously across all related files.

• Search and filter HuggingFace Hub GGUF models inside the llama.cpp UI
• Draft sidecars (dflash, mtp) download alongside the main model
• Pause, resume and cancel apply to all linked files
• The integration can be disabled in settings