| # Use a lightweight Python image | |
| FROM python:3.10-slim | |
| # Set working directory | |
| WORKDIR /app | |
| # ✅ Create necessary cache directories and set correct permissions | |
| RUN mkdir -p /app/.cache/datasets /app/.cache/transformers && chmod -R 777 /app/.cache | |
| # ✅ Copy requirements.txt and install dependencies | |
| COPY requirements.txt /app/requirements.txt | |
| RUN pip install --no-cache-dir -r /app/requirements.txt | |
| # ✅ Copy dataset processing script | |
| COPY process_dataset.py /app/process_dataset.py | |
| # ✅ Run script on container start | |
| CMD ["python", "/app/process_dataset.py"] | |