Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

Large model + vertical domain application Lesser Fullness of Grain open source the first 100 billion-level Chinese financial model in China

2025-04-02 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > IT Information >

Share

Shulou(Shulou.com)12/24 Report--

Today, the application of large model technology in the field is deepening, which continues to promote the transformation and upgrading of the industry. Over the years, Lesser Fullness of Grain has continued to lay out cutting-edge technologies such as large models, promoting the application and promotion of various technologies in the field. Recently, Lesser Fullness of Grain officially opened up China's first 100 billion-level Chinese model-Xuanyuan, showing its scientific and technological strength in the field of large model technology.

Lesser Fullness of Grain focused on large model technology, Xuanyuan's performance surpassed the mainstream open source big model in the market.

It is reported that the Xuanyuan big model is trained on the basis of the Bloom model with 176 billion parameters, and its effect is greatly improved compared with the general model in terms of noun understanding, market comment, data analysis and news understanding, showing obvious domain advantages.

In the task evaluation in the scene, Xuanyuan surpassed the mainstream open source model in the market and won 63.33% of the 150 responses, fully highlighting its significant advantages in the field. In the general ability evaluation, 10.2% of the tasks in Xuanyuan surpassed ChatGPT 3.5, and 61.22% of the tasks were the same, involving 13 main dimensions, such as mathematical calculation, scene writing, logical reasoning, text summary and so on.

In order to improve the ability of Xuanyuan big model to understand domain problems, du Lesser Fullness of Grain used hundreds of billions of tokens Chinese pre-training data sets accumulated in his own business to train the model. The data set covers the expertise of research papers, stocks, banks, and so on. Du Lesser Fullness of Grain said that high-quality data sets after cleaning and tagging not only make it possible to be on a par with ChatGPT in terms of versatility, but also significantly improve the performance of the model in the vertical field.

Lesser Fullness of Grain innovates actively and explores the application of large models in the field.

As a new infrastructure of AI, large model has a wide range of applications in various industries. After the Xuanyuan big model is open source, what is the significance to the organization?

Xu Dongliang believes that the generative model has excellent abilities in content generation and creation, information summary and summary, knowledge understanding and question answering, natural interaction and dialogue, and will be widely used in the scene. At the front desk, the generative model will greatly improve the professional level and service ability of the account manager, greatly reduce the operating cost of the account manager, and make it possible for everyone to have a professional account manager online 24 hours a day. Excellent content generation ability will also lead to a substantial increase in the production capacity of marketing content. In Zhongtai, the generative model has the opportunity to change the way of knowledge acquisition, content creation, conference and communication, code development and testing, which can greatly improve the internal office efficiency, and even lead to the reform of R & D and testing mode. to improve the efficiency of internal operations in an all-round way. In the background, the large model will become the standard of intelligent technology base, greatly reducing the threshold of intelligent technology application, with only a small amount of labeled data or even no need to adjust, intelligent technology can cover a wide range of scenarios.

Relying on Baidu artificial intelligence technology, du Lesser Fullness of Grain has carried out a series of applications based on large models. Taking risk management as an example, du Lesser Fullness of Grain has applied the large language model LLM to the interpretation of Internet text data and credit report. Through the pre-training model constructed by text data and AI algorithm, the credit report can be interpreted into 400,000-dimensional risk variables and better identify the credit risk of small and micro business owners. In May this year, the project won the Wu Wenjun artificial Intelligence Science and Technology Award. In February this year, Baidu's generative dialogue product "Wen Xin Yiyan" (English name: ERNIE Bot), based on Wen Xin big model technology, opened up ecological cooperation, making Lesser Fullness of Grain the first technology company to access.

This time, Lesser Fullness of Grain opened up the first 100 billion-level Chinese model in China, showing a profound technical foundation. For a long time, du Lesser Fullness of Grain has continued to explore in the field of cutting-edge science and technology, constantly using scientific and technological innovation to solve service difficulties and pain points, and helping the development of the real economy. In the future, du Lesser Fullness of Grain will continue to explore the application of technologies such as large models in the scene to bring better service to users.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

IT Information

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report