NEFTune is a simple trick for finetuning language models that can be implemented with only a few lines of code and consistently boosts the resulting model’s performance. Here’s how it works… TL;DR: NEFTune just adds random (uniform) noise to an LLM’s input word… https://t.co/7Vz8AU5Ho4 https://t.co/tTM5waKSsv
— Cameron R. Wolfe, Ph.D. (@cwolferesearch) Nov 8, 2023
from Twitter https://twitter.com/cwolferesearch
November 08, 2023 at 08:29PM
via IFTTT
No comments:
Post a Comment