v0.2 Training: SFT only or SFT+DPO?

#15

by weizechen - opened Sep 6, 2024

Sep 6, 2024

Hi. I've read that the v0.1 documentation mentions SFT+DPO training, while v0.2 only refers to SFT. The alignment handbook also lacks a DPO recipe. Was DPO used for v0.2? Thanks!

loubnabnl

Hugging Face Smol Models Research org Sep 7, 2024

Hi, we only used SFT for v0.2

weizechen

Sep 8, 2024

Thanks for the reply!

weizechen changed discussion status to closed Sep 8, 2024

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

Tap or paste here to upload images

Your need to confirm your account before you can post a new comment.

· Sign up or log in to comment