Dockerfile.gpu

# Dockerfile.gpu - x86_64 with NVIDIA GPU support
FROM python:3.12-slim

LABEL org.opencontainers.image.source=https://github.com/dlaszlo/speech-service

# Set environment
ENV DEBIAN_FRONTEND=noninteractive
ENV PYTHONUNBUFFERED=1
WORKDIR /app

# Install system dependencies
RUN apt-get update && \
    apt-get install -y --no-install-recommends \
        espeak-ng \
        curl \
        ffmpeg \
        && \
    apt-get clean && \
    rm -rf /var/lib/apt/lists/*

# Install uv
COPY --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/uv

# Copy requirements
COPY requirements/ /app/requirements/

# Install GPU dependencies using uv
ENV UV_SYSTEM_PYTHON=1
ENV UV_COMPILE_BYTECODE=1

# We rely on uv/pip to pull standard wheels including CUDA libs where defined in requirements/gpu.txt
RUN uv pip install --system --no-cache -r requirements/gpu.txt

# Add LD_LIBRARY_PATH for NVIDIA libraries (cuDNN, cuBLAS) installed via pip
ENV LD_LIBRARY_PATH="/usr/local/lib/python3.12/site-packages/nvidia/cudnn/lib:/usr/local/lib/python3.12/site-packages/nvidia/cublas/lib${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}"

# Copy source code
COPY . .

# Environment variables
ENV STT_MODEL_NAME="Systran/faster-distil-whisper-small.en"
ENV TTS_MODEL_NAME="hexgrad/Kokoro-82M"
ENV TTS_LANG_CODE="a"
ENV HF_HOME="/data/huggingface"

EXPOSE 8000

HEALTHCHECK --interval=30s --timeout=10s --start-period=60s --retries=3 \
    CMD curl -f http://localhost:8000/health || exit 1

CMD ["uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "8000"]