Pre-training vs. Fine-tuning: A Reproducibility Study on Dense Retrieval Knowledge Acquisition

Open in new window