Chinese data could use this reward to filter?

by Yvonne1111 - opened Mar 3

Mar 3

Shall I use this reward model to filter Chinese data?
I read the Nemotron 340B paper and I find that there are 15% multilingual natural language data in pretrain data.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment