Dockerfile

# Use an official Python runtime as a parent image
FROM python:3.10-slim

# Set environment variables
ENV PYTHONUNBUFFERED=1
ENV HOME=/home/appuser
ENV APP_HOME=/home/appuser/app
ENV PATH="/home/appuser/.local/bin:${PATH}"

# Install system dependencies
RUN apt-get update && apt-get install -y \
    build-essential \
    cmake \
    git \
    wget \
    curl \
    && rm -rf /var/lib/apt/lists/*

# Create a non-root user
RUN useradd -m -u 1000 appuser && \
    mkdir -p $APP_HOME && \
    chown -R appuser:appuser /home/appuser

# Switch to the non-root user
USER appuser
WORKDIR $APP_HOME

# Copy the application files
COPY --chown=appuser:appuser . $APP_HOME/

# Download the model
RUN mkdir -p $APP_HOME/models && \
    wget -O $APP_HOME/models/llama-model.gguf https://huggingface.co./bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/resolve/main/Meta-Llama-3.1-8B-Instruct-IQ2_M.gguf

# Install Python dependencies
RUN pip install --user --no-cache-dir \
    chainlit \
    langchain \
    chromadb \
    sentence-transformers \
    llama-cpp-python \
    --extra-index-url https://download.pytorch.org/whl/cpu

# Create necessary directories with correct permissions
RUN mkdir -p $APP_HOME/.chainlit/files && \
    mkdir -p $APP_HOME/mydb

# Create a chainlit.md file
RUN echo "# Welcome to RAG Chainlit Application! 👋\n\nThis is a Retrieval-Augmented Generation application using Llama.cpp." > $APP_HOME/chainlit.md

# Expose port 8000
EXPOSE 8000

# Start the Chainlit app
CMD ["chainlit", "run", "app.py", "--host", "0.0.0.0", "--port", "8000"]