今日，万兴科技宣布发布国内首个音视频多媒体大模型——“天幕”。该模型聚焦数字创意垂类创作场景，以15亿创作者及100亿本土化高质量音视频数据为基础，运用音视频生成式AI技术，支持全球不同语言，其相关能力已在海外规模化商用。
值得一提的是，万兴“天幕”大模型研发中心将正式落户马栏山。此举标志着我国在音视频大模型领域的研究和应用迈出了重要一步。
此外，万兴科技还与华为云及马投算力达成了三方算力合作。这意味着“天幕”大模型的训练将在国内算力和服务器基础上进行，有助于推动我国音视频大模型技术的发展和应用。
With the release of Tianmu, Wondershare Technology has announced the launch of China’s first audio-video multimedia grand model. The model focuses on creative scenarios in the digital creative industry, based on the massive AV data of 150 million creators and 100 billion locally produced high-quality audio-video clips. Utilizing AI-based audio-video generation technology, it supports multiple languages globally and its relevant capabilities have already been commercially scaled overseas.
It is worth mentioning that the Tianmu grand model research and development center will be officially established in Malan Mountain. This move signifies an important step for China in the field of audio-video grand model research and application.
Furthermore, Wondershare Technology has entered into a trilateral computing power cooperation with Huawei Cloud and Malan Computing Power. This means that the training of the Tianmu grand model will be carried out based on domestic computing power and servers, which is expected to promote the development and application of China’s audio-video grand model technology.
【来源】https://www.tmtpost.com/nictation/6916252.html
Views: 1