scripts/download_notes.md

Automatic Dataset Download

The repo tries to fetch the dataset automatically when you run:

python -m src.cli.run_prepare

Authentication

Use one of these:

• ~/.kaggle/kaggle.json
• KAGGLE_USERNAME and KAGGLE_KEY

Manual fallback

If automatic download is not available in your environment, manually download: US_Accidents_March23.csv

Then place it in: data/raw/US_Accidents_March23.csv