Dockerfile

# Use the official Python image 
FROM python:3.9-slim

# Set the working directory
WORKDIR /app

# Create a writable cache directory for Hugging Face
RUN mkdir -p /workspace/huggingface_cache && \
    chmod -R 777 /workspace/huggingface_cache

# Set environment variable for Hugging Face cache
ENV HF_HOME=/workspace/huggingface_cache

# Create and set permissions for the /.triton directory to avoid permission issues
RUN mkdir -p /root/.triton && chmod -R 777 /root/.triton
ENV TRITON_CACHE_DIR=/root/.triton

# Install system dependencies
RUN apt-get update && apt-get install -y git

# Copy the requirements file into the image
COPY requirements.txt requirements.txt

# Install any needed packages specified in requirements.txt
RUN pip install --no-cache-dir -r requirements.txt

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

# Expose the port your application will run on
EXPOSE 7860

# Command to run your application
CMD ["python", "app.py"]