Update Dockerfile

Dockerfile

@@ -9,15 +9,17 @@ ENV PYTHONDONTWRITEBYTECODE=1 \
     TOKENIZERS_PARALLELISM=false
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
-    gcc g++ make cmake git \
-    libopenblas-dev \
+    git build-essential cmake libopenblas-dev \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /app
 
-# Pre-built llama-cpp-python wheel with OpenBLAS
-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
+# Clone PrismML's llama.cpp fork and build llama-cpp-python from source
+RUN git clone --depth 1 https://github.com/PrismML-Eng/llama-cpp-python.git && \
+    cd llama-cpp-python && \
+    git submodule update --init --recursive && \
+    CMAKE_ARGS="-DGGML_OPENBLAS=ON" pip install --no-cache-dir -e . && \
+    cd .. && rm -rf llama-cpp-python
 
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt