fix: resolve libc.musl error by compiling llama-cpp-python from source in builder stage

Dockerfile

@@ -1,9 +1,14 @@
 # Stage 1: Builder
-# We use a standard Python image and install uv, which is safer and still incredibly fast.
 FROM python:3.10-slim-bookworm AS builder
 WORKDIR /app
 
-# Install uv (The extremely fast Python package installer)
+# Install build tools
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    cmake \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install uv
 RUN pip install uv
 
 # Configure uv
@@ -13,13 +18,12 @@ ENV UV_LINK_MODE=copy
 # Copy requirements
 COPY requirements.txt .
 
-# Create a virtual environment and install dependencies using uv
-# We point to the extra index URL for the pre-built llama-cpp-python wheels
+# Create venv and install dependencies
+# We allow building from source for llama-cpp-python to ensure libc compatibility
 RUN uv venv /app/.venv && \
     uv pip install \
     --no-cache \
     -r requirements.txt \
-    --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu \
     --python /app/.venv
 
 # Stage 2: Final Runtime Image
@@ -29,22 +33,18 @@ WORKDIR /app
 # Copy the virtual environment from the builder stage
 COPY --from=builder /app/.venv /app/.venv
 
-# Set environment variables to use the virtual environment automatically
+# Set environment variables
 ENV PATH="/app/.venv/bin:$PATH"
 
 # Copy the application code
 COPY . .
 
-# Create a non-root user for security (Hugging Face Spaces requirement)
+# Create a non-root user
 RUN useradd -m -u 1000 user
 USER user
 ENV HOME=/home/user
 
-# Expose the application port
 EXPOSE 7860
-
-# Runtime configuration
 ENV MODEL_CTX_SIZE=8192
 
-# Run the application
 CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]