Dockerfile

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

# Set the working directory inside the container
WORKDIR /app

# Copy the file that lists your Python dependencies
COPY ./requirements.txt /app/requirements.txt

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

# This is a key step for performance:
# We run a small script to pre-download the EasyOCR models
# so they are included in the package, avoiding slow server startups.
COPY ./preload_models.py /app/preload_models.py
RUN python /app/preload_models.py

# Copy your application code and simulator into the package
COPY ./app /app/app
COPY ./main.py /app/main.py
COPY ./simulator.html /app/simulator.html

# The command to run your application when it starts
# It listens on all network interfaces (0.0.0.0) on port 7860 for Hugging Face Spaces
CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]