上海人工智能实验室与商汤科技、香港中文大学和复旦大学共同宣布，新一代大语言模型书生·浦语2.0正式开源。该模型的核心理念是回归语言建模的本质，通过提升语料质量及信息密度，实现模型基座语言建模能力的质的提升。
书生·浦语2.0支持200K token的上下文，能够一次性接收并处理约30万汉字的输入内容，准确提取关键信息，实现了在长文本中“大海捞针”的功能。这一模型的发布，标志着我国在人工智能领域的大语言模型研究又迈出了重要的一步。
The release of Scholar·Puyu 2.0, a new-generation large language model, has been officially open-sourced by Shanghai Artificial Intelligence Laboratory, SenseTime Technology, Chinese University of Hong Kong, and Fudan University. The core concept of Scholar·Puyu 2.0 is to return to the essence of language modeling, and to achieve a qualitative leap in the language modeling capabilities of the model’s base through the improvement of the quality and information density of the corpus.
Scholar·Puyu 2.0 supports a context of up to 200K tokens, and can process input content of about 300,000 Chinese characters at once, accurately extracting key information and achieving the function of “fishing for a needle in a haystack” in long texts. The release of this model signifies an important step forward for China’s research in large language models in the field of artificial intelligence.
【来源】https://www.thepaper.cn/newsDetail_forward_26040295
Views: 2