FROM nvidia/cuda:12.1.0-base-ubuntu22.04

# Set working directory
WORKDIR /app

# Install system dependencies
RUN apt-get update && apt-get install -y \
    python3.11 \
    python3-pip \
    ffmpeg \
    libsndfile1 \
    sox \
    libsox-dev \
    libsox-fmt-all \
    && rm -rf /var/lib/apt/lists/*

# Copy requirements
COPY requirements.txt .

# Upgrade pip to avoid dependency resolution issues
RUN pip3 install --upgrade pip

# Install dependencies for sox package (required by NeMo) in correct order
RUN pip3 install --no-cache-dir numpy==2.2.2 typing-extensions

# Install Python dependencies with legacy resolver (NeMo has complex dependencies)
RUN pip3 install --no-cache-dir --use-deprecated=legacy-resolver -r requirements.txt

# Copy application code
COPY . .

# Create models directory
RUN mkdir -p /models

# Expose port
EXPOSE 8000

# Set environment variables
ENV PYTHONUNBUFFERED=1
ENV CUDA_VISIBLE_DEVICES=0
ENV LD_LIBRARY_PATH=/usr/local/lib/python3.11/dist-packages/nvidia/cudnn/lib:${LD_LIBRARY_PATH}

# Run the server
CMD ["uvicorn", "stt_server:app", "--host", "0.0.0.0", "--port", "8000", "--log-level", "info"]