FROM nvidia/cuda:12.4.1-devel-ubuntu22.04

ENV DEBIAN_FRONTEND=noninteractive
ENV NVIDIA_VISIBLE_DEVICES=all
ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility

# No deadsnakes PPA needed. Native Python 3.10 works perfectly.
RUN apt-get update && apt-get install -y \
    python3 python3-dev python3-pip curl git libsndfile1 ffmpeg sox libsox-dev ninja-build && \
    rm -rf /var/lib/apt/lists/*

WORKDIR /app

# Install Torch and core packages
RUN python3 -m pip install --no-cache-dir torch==2.6.0 torchaudio --index-url https://download.pytorch.org/whl/cu124

# 1. Install foundational build tools and numpy first
RUN python3 -m pip install --no-cache-dir numpy setuptools wheel ninja packaging psutil

# 2. Install the rest of the stack that relies on numpy being present
RUN python3 -m pip install --no-cache-dir fastapi uvicorn soundfile librosa transformers==4.57.3 accelerate sox onnxruntime

# Force ABI compatibility for the C++ compiler
ENV _GLIBCXX_USE_CXX11_ABI=0
ENV MAX_JOBS=8

# Install flash-attn
RUN python3 -m pip install --no-cache-dir flash-attn --no-build-isolation

COPY swarm-control/indra-tts-server/tts-server.py .

EXPOSE 8002
CMD ["python3", "tts-server.py"]