VLFeedback VLFeedback and Silkie

A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment

Lei Li*, Zhihui Xie*, Mukai Li,
Shunian Chen, Peiyi Wang, Liang Chen,
Yazheng Yang, Benyou Wang, Lingpeng Kong, Qi Liu

▶ The University of Hong Kong
▶ The Chinese University of Hong Kong, Shenzhen
▶ Peking University
  *Equal Contribution
arXiv Code Dataset (VLFeedback) Model

Accepted by EMNLP 2024 Main Conference

VLFeedback is the first open-sourced GPT-4V annotated vision-language preference dataset, covering 80k instructions sampled from various sources with responses decoded from 12 large language vision models such as GPT-4V, LLaVA-series and Qwen-VL.
Based on Qwen-VL-Chat, we present Silkie, by performing DPO on our VLFeedback. Compared with the original model, Silkile achieves 6.9% and 9.5% relative improvement on the MME benchmark regarding the perception and cognition capabilities, respectively. Besides, Silkie sets a new state-of-the-art score of 3.02 on MMHal-Bench regarding hallucination evaluation.

Abstract

As large vision-language models (LVLMs) evolve rapidly, the demand for high-quality and diverse data to align these models becomes increasingly crucial. However, the creation of such data with human supervision proves costly and time-intensive. In this paper, we investigate the efficacy of AI feedback to scale supervision for aligning LVLMs. We introduce VLFeedback, the first large-scale vision-language feedback dataset, comprising over 82K multi-modal instructions and comprehensive rationales generated by off-the-shelf models without human annotations. To evaluate the effectiveness of AI feedback for vision-language alignment, we train Silkie, an LVLM fine-tuned via direct preference optimization on VLFeedback. Silkie showcases exceptional performance regarding helpfulness, visual faithfulness, and safety metrics. It outperforms its base model by 6.9% and 9.5% in perception and cognition tasks, reduces hallucination issues on MMHal-Bench, and exhibits enhanced resilience against red-teaming attacks. Furthermore, our analysis underscores the advantage of AI feedback, particularly in fostering preference diversity to deliver more comprehensive improvements.

Multimodal Instructions and AI Preference Data

We sample multi-modal instructions from various souces, covering different capabilities of LVLMs. We further build a model pool consisting of 12 LVLMs. .


We further use GPT-4V as the annoator to assess the quality of each response regarding helpfulessn, visual faithfulness, and ethical considerations.



(Left) The overall score distribution of three aspects. (Right) The comparison of models in our pool.

Silkie: A Better Aligned LVLM

We improve Qwen-VL-Chat by performing DPO on our VLFeedback, using the efficient LoRA tuning method. After DPO training, the resulting model Silkie achieves promising results compared with other models with similar-sized LLMs as the backbone.


(Left) In-depth analysis on the MME benchmark for the performance improvements. Our VLFeedback dataset brings clearer gains in OCR recognition and code reasoning tasks.
(Right) Relative performance improvement by performing DPO with RLHF-V preference data and a subset of our VLFeedback dataset. Our GPT-4V annotated preference dataset brings more consistent improvements on four benchmarks.

Qualitative Examples

Our Silkie locates the wooden stools with a red flower without giving misleading assertions (Left), and correctly answers the scientific-related question (Right), exhibiting better perception and cognition capabilities.

On a challenging query asking the model to generate a report for the diagram of weather forecast process, Silkie generates a well-structured report satisfying the word requirement.

BibTeX


  @article{2023vlfeedback,
    author      = {Lei Li and Zhihui Xie and Mukai Li and Shunian Chen and Peiyi Wang and Liang Chen and  Yazheng Yang and  Benyou Wang and  Lingpeng Kong},
    title       = {Silkie: Preference Distillation for Large Visual Language Models},
    publisher   = {arXiv:2312.10665},
    year        = {2023}
  }

Acknowledgement

This website is adapted from Nerfies and LLaVA-RLHF, licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. We thank the authors of the multi-modal instruction tuning datasets and open-source projects, including LLaVA, LLaVA-RLHF and Qwen-VL. We would thank Runxin Xu for his great help on the project.

Usage and License Notices: The data, code and checkpoint is intended and licensed for research use only. They are also restricted to uses that follow the license agreement of Qwen-VL and GPT-4. The dataset is CC BY NC 4.0 (allowing only non-commercial use) and models trained using the dataset should not be used outside of research purposes.

Related Links: [LLaVA] [LLaVA-RLHF] [Qwen-VL]