MultiModal-GPT: A Vision and Language Model for Dialogue with Humans

TL;DR

MultiModal-GPT is a parameter-efficient vision and language model designed for multi-round human dialogue, capable of understanding and following diverse instructions through multi-modality instruction tuning and joint training.

Contribution

It introduces a multi-modal instruction tuning approach with vision and language data, and demonstrates improved dialogue performance via joint training with language-only data.

Findings

Effective multi-round dialogue with humans demonstrated

Joint training with language-only data enhances dialogue quality

Model responds accurately to diverse vision-language instructions

Abstract

We present a vision and language model named MultiModal-GPT to conduct multi-round dialogue with humans. MultiModal-GPT can follow various instructions from humans, such as generating a detailed caption, counting the number of interested objects, and answering general questions from users. MultiModal-GPT is parameter-efficiently fine-tuned from OpenFlamingo, with Low-rank Adapter (LoRA) added both in the cross-attention part and the self-attention part of the language model. We first construct instruction templates with vision and language data for multi-modality instruction tuning to make the model understand and follow human instructions. We find the quality of training data is vital for the dialogue performance, where few data containing short answers can lead the model to respond shortly to any instructions. To further enhance the ability to chat with humans of the MultiModal-GPT, we utilize language-only instruction-following data to train the MultiModal-GPT jointly. The joint training of language-only and visual-language instructions with the \emph{same} instruction template effectively improves dialogue performance. Various demos show the ability of continuous dialogue of MultiModal-GPT with humans. Code, dataset, and demo are at https://github.com/open-mmlab/Multimodal-GPT