[HN Gopher] If you are using LLM RAG - you should be doing RAFT
___________________________________________________________________
If you are using LLM RAG - you should be doing RAFT
Author : shishirpatil
Score : 24 points
Date : 2024-03-19 18:31 UTC (4 hours ago)
(HTM) web link (techcommunity.microsoft.com)
(TXT) w3m dump (techcommunity.microsoft.com)
| tianjunz wrote:
| This is a three-way collaboration between Berkeley AI, Microsoft
| Azure, and Meta AI! RAFT involves the concept of domain-specific
| RAG, which represents a more focused and growingly favored area
| compared to the broader concept of the general open-book exam. In
| such exams, the domain in which the LLM will be evaluated is
| known in advance and used for inference. The LLM is capable of
| addressing prompts by leveraging any and all information from
| this particular domain, on which it has been specifically fine-
| tuned.
|
| Blogs: https://gorilla.cs.berkeley.edu/blogs/9_raft.html
| petervandijck wrote:
| It's "Retrieval Augmented Fine Tuning". The related blog post is
| interesting: https://gorilla.cs.berkeley.edu/blogs/9_raft.html
| jondwillis wrote:
| Page me when this process is at least partially automated and
| continuous.
| jnwatson wrote:
| Does any of this matter when you have 5 million token input and
| can just shove everything into the input?
| jscheel wrote:
| Yes, because 5 million token input is slow, expensive, and
| error-prone.
| catchnear4321 wrote:
| > They hypothesized that a student who studies the textbooks
| before the open-book exam was likely to perform better than a
| student who studies the textbook.
|
| interesting hypothesis, but probably not what was meant.
|
| "editor" is shaping up to be the hot new job for fleshy
| consciousnesses by the early thirties.
| matt3D wrote:
| Am I missing the point or is this not how everyone was doing RAG
| before?
|
| I've had my much greater success doing the RAG process on a fine
| tuned model.
| dragonwriter wrote:
| > Am I missing the point or is this not how everyone was doing
| RAG before?
|
| No, especially people wrapping RAG frameworks around models
| they don't have access to fine tune (e.g., GPT-4 if you aren't
| OpenAI and/or Microsoft.)
|
| Edit: Of course, further evidence that this method is useful
| doesn't help people in that condition, though.
___________________________________________________________________
(page generated 2024-03-19 23:01 UTC)