ALLaVA-4V

Name: ALLaVA-4V
Creator: OpenDataLab
License: 暂无描述

OpenXLab2026-04-18 收录

下载链接：

https://openxlab.org.cn/datasets/OpenDataLab/ALLaVA-4V

下载链接

链接失效反馈

官方服务：

资源简介：

ALLaVA-4V dataset is created by leveraging GPT-4V’s ability to generate detailed captions

提供机构：

OpenDataLab

创建时间：

2024-04-30

ALLaVA-4V

## 📚 ALLaVA-4V Data ### Generation Pipeline --> * LAION We leverage the superb GPT-4V to

魔搭社区2026-04-30 更新150

ALLaVA-4V-Arabic

## ALLaVA-4V for Arabic This is the Arabic version of the ALLaVA-4V data. We have translated the ALLaVA-4V data into Arabic through ChatGPT and instructed ChatGPT not to translate content related to O

魔搭社区2025-11-02 更新40

FreedomIntelligence/ALLaVA-4V-Chinese

--- license: apache-2.0 task_categories: - question-answering - text-generation language: - zh tags: - GPT-4V - LVLM - Vision - Language size_categories: - 1M<n<10M configs: - config_name: allava_l

Hugging Face2024-04-29 更新870

FreedomIntelligence/ALLaVA-4V

ALLaVA-4V数据集是一个包含图像描述和复杂推理问答对的多模态数据集，主要使用GPT-4V和GPT-4-Turbo模型生成数据。数据集分为多个子集，包括ALLaVA-Caption-LAION-4V、ALLaVA-Caption-VFLAN-4V、ALLaVA-Instruct-LAION-4V、ALLaVA-Instruct-VFLAN-4V和Evol-Instruct-GPT4-Turbo

Hugging Face2025-06-08 更新620

ALLaVA-4V-Chinese

## ALLaVA-4V for Chinese This is the Chinese version of the ALLaVA-4V data. We have translated the ALLaVA-4V data into Chinese through ChatGPT and instructed ChatGPT not to translate content related t

魔搭社区2025-12-05 更新1070