fix: use pre-built llama-cpp-python wheel to avoid build timeout

Source compilation of llama.cpp takes 10+ minutes and times out.
The abetlen wheel index provides pre-built manylinux wheels for CPU.

requirements.txt

@@ -1,2 +1,3 @@
+--extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu
 llama-cpp-python>=0.3.0
 huggingface-hub>=0.26.0