Getting LLaMA to produce structured outputs (e.g., JSON) is a challenge. @evanqjones + @GrantSlatton work on grammar-based sampling is a cool approach: supply a grammar file to guide / constrain sampling. Thanks @deepsense_ai for adding to @LangChainAI llamacpp integration ... https://t.co/xrKRuwijKo
— Lance Martin (@RLanceMartin) Aug 28, 2023
from Twitter https://twitter.com/RLanceMartin
August 28, 2023 at 11:41AM
via IFTTT
No comments:
Post a Comment