News for hackers. Stuff that matters.

A comment on LLMLingua: Compressing Prompts for Faster Inferencing.

Re:LLMLingua: Compressing Prompts for Faster Inferencing by baq on Tuesday December 19, 2023 @03:41AM (#38693267)
Redundancy is resiliency - wonder if there’s still enough error correction in the compressed language?
[ Parent | on HN ]
Re:LLMLingua: Compressing Prompts for Faster Inferencing by samus on Tuesday December 19, 2023 @07:09AM (#38694589)
Error connection is not really required here since there is no lossy communication medium.
[ Parent | on HN ]
Re:LLMLingua: Compressing Prompts for Faster Inferencing by Kevin09210 on Tuesday December 19, 2023 @04:05PM (#38701534)
Until your prompts looks like the following:

<Do something> with the following text:

<some text>

[ Parent | on HN ]