Dockerfile

# Use a lightweight official Python image
FROM python:3.11-slim

# Install system dependencies required for C++ binaries, file operations, and curl
RUN apt-get update && apt-get install -y --no-install-recommends \
    build-essential \
    curl \
    libgomp1 \
    && rm -rf /var/lib/apt/lists/*

# Set up a working directory inside the container
WORKDIR /code

# Copy requirement configuration file first to leverage Docker layer caching
COPY ./requirements.txt /code/requirements.txt

# Install python dependencies
RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt

# Copy the rest of your application code into the container
COPY . .

# CRITICAL: Grant executable permissions to your C++ inference binary
RUN chmod +x /code/fastvlm_server

# Create a secure temporary directory for permissions compliance on Hugging Face
RUN mkdir -p /tmp && chmod 777 /tmp
ENV TMPDIR=/tmp

# Hugging Face Spaces expect containers to listen on port 7860
EXPOSE 7860

RUN chmod +x /code/start.sh

# Run the orchestration script
CMD ["./start.sh"]