美国国家工程院外籍院士沈向洋在近期的一次演讲中指出，随着人工智能技术的不断进步，通用大模型将成为未来科技发展的重要趋势。他预测，这些模型将在各个垂直行业中发挥巨大作用，并推动整个行业的发展。沈向洋强调，为了达到通用模型的性能要求，模型规模将需要达到万卡级别，参数数量更将达到上万亿。
沈向洋进一步提出，未来的发展方向可能是个性化的个人大模型。他认为，将个性化参数与云计算和端设备相结合，将开创出极具价值的新局面。这种结合不仅能够为用户提供更加精准和个性化的服务，还能够有效提升模型的性能和效率。
英文标题： Shen Xiangyang on the Era of Universal Large Models
英文关键词： Universal Models, Personalized Parameters, Cloud Computing Development
英文新闻内容：
Shen Xiangyang, a foreign member of the US National Academy of Engineering, recently delivered a speech discussing the trend of universal large models in the future development of artificial intelligence. He predicts that these models will have a significant impact on various vertical industries, driving their growth. Shen emphasizes that to achieve the performance of universal models, the scale of the models will require reaching the level of ten thousand cards, with parameter numbers reaching trillions.
Shen further suggests that the future development direction may be personalized large models for individuals. He believes that combining personalized parameters with cloud computing and edge devices can open up a new and valuable situation. Such a combination can not only provide more accurate and personalized services for users but also effectively improve the performance and efficiency of the models.
【来源】https://www.cls.cn/detail/1628008
Views: 1