# Hugging Face — Preference Optimization for Vision Language Models

- Company: Hugging Face (huggingface.co)
- Announced: 2024-07-10T00:00:00+00:00
- Category: developer-tool-release
- Subject: Platform
- Models affected: Idefics2-8b, Llava 1.5, PaliGemma
- Source: https://huggingface.co/blog/dpo_vlm
- Record: https://forck.live/items/1858-preference-optimization-for-vision-language-models

Hugging Face announces that the TRL library now supports direct preference optimization (DPO) for vision language models (VLMs), including Idefics2-8b, Llava 1.5, and PaliGemma.

## Evidence

Verbatim from https://huggingface.co/blog/dpo_vlm:

> We are excited to announce that the TRL library now supports direct preference optimization (DPO) for VLMs.

---

Record: https://forck.live/items/1858-preference-optimization-for-vision-language-models
Catalogue: https://forck.live/llms.txt
Feed: https://forck.live/feed.md