近日，上海人工智能实验室与商汤科技联合香港中文大学和复旦大学共同发布新一代大语言模型书生·浦语2.0。该模型专注于回归语言建模的本质，通过提高语料质量和信息密度，实现了模型基座语言建模能力的显著提升。书生·浦语2.0支持200K token的上下文处理能力，能够在一次性接收并处理约30万汉字输入内容的同时，准确提取关键信息，从而在长文本中实现“大海捞针”。
Title: Shanghai AI Lab Releases Bookworm Puyu 2.0
Keywords: Shanghai AI Lab, Bookworm Puyu 2.0, Language Modeling
News content:
Recently, the Shanghai AI Lab, in collaboration with SenseTime and universities including the Chinese University of Hong Kong and Fudan University, has unveiled a new generation of large language model, Bookworm Puyu 2.0. Focused on the essence of language modeling, this model aims to improve the quality and density of its corpus to significantly enhance the language modeling capabilities of its foundation. Bookworm Puyu 2.0 supports processing up to 200K tokens of context, enabling it to accurately extract key information from approximately 300,000 characters of input content at once, effectively “finding a needle in a haystack” within long texts.
【来源】https://www.thepaper.cn/newsDetail_forward_26040295
Views: 1