feat: add contextualize_chunks() for LLM-prefixed chunk embeddings f181551 benroshan commited on Jun 20