chore: trigger Hugging Face Space rebuild

fix(docker): limit parallel compile to avoid OOM (exit 137) on HF build

fix: switch to Docker SDK so llama-cpp-python compiles against glibc

fix: compile llama-cpp-python from source via packages.txt cmake

fix: switch to --find-links GitHub releases for llama-cpp-python wheel

fix: enable GGUF generation in Space via pre-built llama-cpp-python CPU wheel

Reinitialize Space repo with latest code