Dockerfile.gpu

# OpenPronounce web app (FastAPI + UI), CUDA build.
#
#   docker build -f Dockerfile.gpu -t openpronounce:gpu .
#   docker run --gpus all -p 8000:8000 openpronounce:gpu
#
# Requires the NVIDIA Container Toolkit on the host. Models are downloaded at build
# time; inference runs on the GPU when one is visible (OPENPRONOUNCE_DEVICE=cpu to force CPU).
FROM nvidia/cuda:12.4.1-cudnn-runtime-ubuntu22.04

ENV PYTHONDONTWRITEBYTECODE=1 \
    PYTHONUNBUFFERED=1 \
    PIP_NO_CACHE_DIR=1 \
    HF_HOME=/models \
    PORT=8000 \
    DEBIAN_FRONTEND=noninteractive

RUN apt-get update \
    && apt-get install -y --no-install-recommends python3 python3-pip python3-venv ffmpeg espeak-ng libsndfile1 \
    && rm -rf /var/lib/apt/lists/* \
    && ln -sf /usr/bin/python3 /usr/local/bin/python

WORKDIR /app

# The bundled pip is too old to resolve the app extras (ResolutionImpossible on
# uvicorn[standard], see issue #11).
RUN python -m pip install --upgrade pip

RUN pip install torch --index-url https://download.pytorch.org/whl/cu124

COPY pyproject.toml README.md ./
COPY openpronounce ./openpronounce
RUN pip install ".[app]"

RUN python -c "from openpronounce.speech import _load_models; _load_models()" \
    && python -c "from openpronounce.phones import _load_model; _load_model()"

COPY server.py ./
COPY templates ./templates
COPY static ./static

EXPOSE 8000
CMD uvicorn server:app --host 0.0.0.0 --port ${PORT}