←back to thread

Guide to Fine-Tuning LLMs

(arxiv.org)
145 points ignoramous | 4 comments | 22 Oct 24 04:41 UTC | HN request time: 0.491s | source
1. anothername12 ◴[22 Oct 24 06:33 UTC] No.41911748[source]▶
Well, it sucks that we’re at the “best practices” phase already
replies(3): >>41911886 #>>41912332 #>>41912970 #
2. p1esk ◴[22 Oct 24 07:02 UTC] No.41911886[source]▶
It sucks that we’re still at “best practices” phase. We’ve been in this phase for the last three decades [1], and I really hope we enter “good theory” phase soon.

[1] https://cseweb.ucsd.edu/classes/wi08/cse253/Handouts/lecun-9...

3. kleiba ◴[22 Oct 24 08:32 UTC] No.41912332[source]▶
Why is that?
4. make3 ◴[22 Oct 24 10:37 UTC] No.41912970[source]▶
there's likely still an infinite amount of things to figure out, transformers haven't been out for 10 years yet