Can LLMs themselves do retrieval/reranking? One of the main challenges includes being able to stuff enough document chunks into the LLM context window so it can decide which ones are most relevant. Had a short talk at the @ollama meetup last night (thanks @jmorgan and co. for… https://t.co/pVrlnvVOx0 https://t.co/c5z127PPZK
— Jerry Liu (@jerryjliu0) Jan 24, 2024
from Twitter https://twitter.com/jerryjliu0
January 24, 2024 at 08:08PM
via IFTTT
No comments:
Post a Comment