Frequently Asked Questions

Why does the first dataset load take longer?

If download=True, required files are fetched from Hugging Face and cached under data_path. Subsequent runs reuse the local cache.

How do I load just a small subset quickly?

Pick a single run key from pnpl.datasets.libribrain2025.constants.RUN_KEYS and pass it via include_run_keys=[...].

How do I standardize data?

Datasets standardize by default using per‑run channel stats. Disable with standardize=False or provide channel_means/channel_stds explicitly.

How do I cite PNPL?

See the Citation page (coming soon). For now, please link to the project repository.