全融合技术报告

OmniFusion Technical Report

April 9, 2024

作者: Elizaveta Goncharova, Anton Razzhigaev, Matvey Mikhalchuk, Maxim Kurkin, Irina Abdullaeva, Matvey Skripkin, Ivan Oseledets, Denis Dimitrov, Andrey Kuznetsov

cs.AI

摘要

去年，多模态架构在基于人工智能的方法和解决方案中掀起了一场革命，扩展了大型语言模型（LLM）的能力。我们提出了一种基于预训练LLM和用于视觉模态的适配器的OmniFusion模型。我们评估并比较了几种架构设计原则，以实现更好的文本和视觉数据耦合：MLP和Transformer适配器，各种基于CLIP ViT的编码器（如SigLIP、InternVIT等）及其融合方法，图像编码方法（整个图像或瓷砖编码）以及两个7B的LLM（专有和开源的Mistral）。在8个视觉-语言基准测试上进行的实验显示，与开源的LLaVA类似解决方案（VizWiz、Pope、MM-Vet、ScienceQA、MMBench、TextVQA、VQAv2、MMMU）相比，最佳OmniFusion设置在不同VQA任务方面获得了最高分。我们还提出了多种情境，OmniFusion在不同领域提供了高度详细的答案：家政、观光、文化、医学、手写和扫描方程式识别等。基于Mistral的OmniFusion模型是一个开源解决方案，其权重、训练和推断脚本可在https://github.com/AIRI-Institute/OmniFusion 上获取。

English

Last year, multimodal architectures served up a revolution in AI-based approaches and solutions, extending the capabilities of large language models (LLM). We propose an OmniFusion model based on a pretrained LLM and adapters for visual modality. We evaluated and compared several architecture design principles for better text and visual data coupling: MLP and transformer adapters, various CLIP ViT-based encoders (SigLIP, InternVIT, etc.), and their fusing approach, image encoding method (whole image or tiles encoding) and two 7B LLMs (the proprietary one and open-source Mistral). Experiments on 8 visual-language benchmarks show the top score for the best OmniFusion setup in terms of different VQA tasks in comparison with open-source LLaVA-like solutions: VizWiz, Pope, MM-Vet, ScienceQA, MMBench, TextVQA, VQAv2, MMMU. We also propose a variety of situations, where OmniFusion provides highly-detailed answers in different domains: housekeeping, sightseeing, culture, medicine, handwritten and scanned equations recognition, etc. Mistral-based OmniFusion model is an open-source solution with weights, training and inference scripts available at https://github.com/AIRI-Institute/OmniFusion.