# Use NVIDIA CUDA base image with Python 3.12 (CUDA 12.8 for RTX 5080)
FROM nvidia/cuda:12.8.0-base-ubuntu24.04

# Set environment variables
ENV DEBIAN_FRONTEND=noninteractive \
    PYTHONUNBUFFERED=1 \
    PYTHONDONTWRITEBYTECODE=1 \
    PIP_NO_CACHE_DIR=1 \
    CUDA_VISIBLE_DEVICES=0

# Configure apt to use Tsinghua mirror (清华源)
RUN sed -i 's@//archive.ubuntu.com@//mirrors.tuna.tsinghua.edu.cn@g' /etc/apt/sources.list.d/ubuntu.sources && \
    sed -i 's@//security.ubuntu.com@//mirrors.tuna.tsinghua.edu.cn@g' /etc/apt/sources.list.d/ubuntu.sources

# Install Python and system dependencies (Ubuntu 24.04 uses Python 3.12)
RUN apt-get update && apt-get install -y \
    python3 \
    python3-pip \
    python3-venv \
    git \
    libglib2.0-0 \
    libsm6 \
    libxext6 \
    libxrender-dev \
    libgomp1 \
    wget \
    && rm -rf /var/lib/apt/lists/*

# Create symlink for python command
RUN ln -sf /usr/bin/python3 /usr/bin/python

# Configure pip to use Tsinghua mirror (清华源) and allow system-wide installs
RUN python3 -m pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple && \
    python3 -m pip config set global.break-system-packages true

# Upgrade pip (ignore system-installed packages)
RUN pip install --upgrade --ignore-installed pip setuptools wheel

# Set working directory
WORKDIR /app

# Copy project files
COPY . /app/

# Install PyTorch with CUDA support first (cu124 is compatible with CUDA 12.8)
# Note: PyTorch uses official mirror as Tsinghua doesn't host CUDA builds
RUN pip install torch torchvision

# Install the package and dependencies
# Set version manually since .git is excluded by .dockerignore
ENV SETUPTOOLS_SCM_PRETEND_VERSION=1.0.0
RUN pip install -e .

# Install additional dependencies for server
RUN pip install requests

# Expose port for Ray Serve
EXPOSE 8001

# Create cache directory for models
RUN mkdir -p /root/.cache/huggingface/hub

# Health check
HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
    CMD python3 -c "import requests; requests.get('http://localhost:8001/', timeout=5)" || exit 1

# Default command to start the server (port 8001)
CMD ["texteller", "launch", "-p", "8001"]