One-stop pipeline for building IR datasets from PDFs and fine-tuning embedding models. - View it on GitHub