Dockerfile

FROM nvidia/cuda:12.3.2-cudnn9-devel-ubuntu22.04

# Set environment variables
ENV PYTHONUNBUFFERED=1 \
    DEBIAN_FRONTEND=noninteractive \
    CUDA_HOME=/usr/local/cuda \
    PATH=/usr/local/cuda/bin:$PATH \
    LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH \
    NVIDIA_VISIBLE_DEVICES=all \
    NVIDIA_DRIVER_CAPABILITIES=compute,utility \
    HF_HOME=/app/models \
    NUMBA_CACHE_DIR=/tmp/numba_cache

# Install system dependencies
RUN apt-get update && apt-get install -y --no-install-recommends \
    python3 \
    python3-pip \
    python3-dev \
    build-essential \
    git \
    ffmpeg \
    libsndfile1 \
    curl \
    && rm -rf /var/lib/apt/lists/*

# Upgrade pip and install build tools
RUN python3 -m pip install --upgrade pip setuptools wheel uv

WORKDIR /app

# Create Numba cache directory
RUN mkdir -p /tmp/numba_cache && \
    chown nobody:nogroup /tmp/numba_cache && \
    chmod 700 /tmp/numba_cache

COPY requirements.txt .

# Install other requirements
RUN python3 -m uv pip install --no-cache-dir -r requirements.txt --prerelease=allow
RUN python3 -m uv pip install --no-build-isolation flash-attn

COPY . .

EXPOSE 8000

CMD ["python3", "server.py"]