Use prebuilt llama-cpp-python wheel for faster builds

- Install from Luigi's prebuilt wheel (0.3.22-cp310-cp310-linux_x86_64)
- Remove build tools (build-essential, cmake) from Dockerfile
- Remove llama-cpp-python from requirements.txt
- Build time: ~4 minutes → ~3 seconds (98% faster)

Dockerfile

@@ -2,22 +2,18 @@ FROM python:3.10-slim
 
 WORKDIR /app
 
-# Install system dependencies (build tools needed for llama-cpp-python)
+# Install minimal system dependencies (OpenBLAS for the prebuilt wheel)
 RUN apt-get update && apt-get install -y \
-    build-essential \
-    cmake \
-    python3-dev \
+    libopenblas-dev \
     libopencc-dev \
     && rm -rf /var/lib/apt/lists/*
 
-# Copy requirements first for better caching
-COPY requirements.txt .
-
-# Install llama-cpp-python with CPU-only support (no AVX requirements for broader compatibility)
-RUN CMAKE_ARGS="-DLLAMA_AVX=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF -DLLAMA_F16C=OFF" \
-    pip install --no-cache-dir llama-cpp-python
+# Install llama-cpp-python from prebuilt wheel (FAST - no build needed!)
+RUN pip install --no-cache-dir \
+    https://huggingface.co/Luigi/llama-cpp-python-wheels-hf-spaces-free-cpu/resolve/main/llama_cpp_python-0.3.22-cp310-cp310-linux_x86_64.whl
 
-# Install remaining requirements
+# Copy and install other requirements
+COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
 # Copy application files

requirements.txt

@@ -1,4 +1,3 @@
 gradio>=5.0.0
 opencc-python-reimplemented>=0.1.7
 huggingface-hub>=0.23.0
-llama-cpp-python>=0.3.0