by Labo333 on 12/14/23, 3:13 PM with 20 comments
by AceJohnny2 on 12/14/23, 5:58 PM
https://en.wikipedia.org/wiki/LoRa
Instead it's about LoRA, note the capitalized last A, or Low-Rank Adaptation, a method for tuning LLMs.
by taneq on 12/14/23, 4:11 PM
‘Conventional’ (if that means anything in a field 10 minutes old) wisdom is “fine tune to add knowledge, LoRA to adjust presentation” - could you comment on your experiences with this?
by dtks on 12/14/23, 8:50 PM
https://github.com/predibase/lorax
Does a similar thing and seems more active.
by SubiculumCode on 12/14/23, 4:20 PM