Dockerfile

# Use a modern, slim Python base image
FROM python:3.11-slim

# Set the working directory
WORKDIR /app

# Install the essential build tools (C++ compiler, CMake) and git
RUN apt-get update && apt-get install -y --no-install-recommends \
    g++ \
    cmake \
    git \
    build-essential

# This is the critical step:
# Install llama-cpp-python while forcing a CPU-optimized build using OpenBLAS.
# This avoids the memory errors and stalling seen with the default build process.
RUN CMAKE_ARGS="-DLLAMA_BLAS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS" \
    pip install llama-cpp-python --no-cache-dir

# Now, install the other dependencies from requirements.txt
COPY requirements.txt .
RUN pip install -r requirements.txt --no-cache-dir

# Copy the rest of your application code
COPY . .

# Expose the port Gradio runs on
EXPOSE 7860

# The command to run your application
CMD ["python", "app.py"]