近日，我国知名人工智能公司零一万物发布了全新的多模态大模型Yi Vision Language（Yi-VL），并正式面向全球开源。该模型基于Yi语言模型开发，包含Yi-VL-34B和Yi-VL-6B两个版本，以其卓越的图文理解和对话生成能力，在英文数据集MMMU和中文数据集CMMMU上取得了领先成绩。
Yi-VL模型的发布，展示了零一万物在人工智能领域的强大实力，进一步推动了多模态大模型在全球范围内的研究和发展。据悉，Yi-VL模型不仅在图文理解和对话生成方面表现出色，还具备强大的跨学科任务处理能力，为人工智能应用的拓展提供了更多可能性。
作为一款开源模型，Yi-VL的发布也将促进全球人工智能社区的交流与合作，为研究人员和开发者提供了宝贵的资源和工具。未来，随着Yi-VL模型的不断优化和升级，我们有理由相信，人工智能技术将会在更多领域取得突破性的进展。
英文翻译：
Recently, the well-known artificial intelligence company Zero One released a brand new multimodal large model, Yi Vision Language (Yi-VL), and officially opened it up to the global community. The model, based on the Yi language model, includes two versions: Yi-VL-34B and Yi-VL-6B. With its exceptional ability in image and text understanding and dialogue generation, Yi-VL has achieved leading results on the English dataset MMMU and the Chinese dataset CMMMU.
The release of Yi-VL model demonstrates Zero One’s strong strength in the field of artificial intelligence, further promoting the research and development of multimodal large models globally. It is learned that Yi-VL model not only excels in image and text understanding and dialogue generation but also possesses strong cross-discipline task handling capabilities, providing more possibilities for the expansion of artificial intelligence applications.
As an open-source model, the release of Yi-VL will also promote communication and cooperation among the global artificial intelligence community, providing researchers and developers with valuable resources and tools. In the future, with the continuous optimization and upgrading of the Yi-VL model, we have every reason to believe that artificial intelligence technology will make breakthrough progress in more fields.
【来源】https://www.jiqizhixin.com/articles/2024-01-22-10
Views: 1