i think a better approach might be to addly ask GPT-4 to extract a short key phrase from the chunk to base its Q/A on, and then train the model on (q, short phrase, full chunk, answer). then either hallucinate both or just do retrieval based on the short phrase alone
in reply to: 1723945812663664964
same thread: 1723945797945844095 1723945800995381617 1723945803792666894 1723945808041533708 1723945812663664964 1723945818569273350 1723945821782053321
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.