Shulou(Shulou.com)11/24 Report--

Continuous innovation of large model technology will further contribute to the transformation and upgrading of the financial industry. However, there are still many challenges in how to give full play to the advantages of large models in the financial field and speed up the application of large model technology to landing finance. In this regard, du Lesser Fullness of Grain focuses on the large model, uses innovative technology to deal with some practical difficulties in the financial field of the landing of the large model, and promotes the innovation and development of the financial industry.

Du Lesser Fullness of Grain focuses on the big model and speeds up the pace of technological innovation.

On August 23, Sinorama-du Lesser Fullness of Grain Financial Technology Laboratory of Peking University held the "Peking University Guanghua-degree Lesser Fullness of Grain Financial Model Technology and Application Forum". Professor Liu Xiaolei, Director of Guanghua Finance Department of Peking University, Xu Dongliang, du Lesser Fullness of Grain, CTO, and Sun Maosong, executive vice president of artificial Intelligence Research Institute of Tsinghua University, attended the seminar. At the meeting, the technical difficulties of the large model itself, the impact and application of the large model on the financial industry have become topics of common concern in the industry-university-research community.

Du Xiaoman CTO Xu Dongliang

Xu Dongliang said that the financial industry is a high-value industry with a good digital foundation, highly dependent on data and technology, and is a high-potential scenario for the landing application of large models. For small and medium-sized financial institutions, in the wave of large models, they also have the opportunity to speed up their own digitization and intelligence process and bridge the digital gap through applied innovation.

The big open source model Xuanyuan has attracted hundreds of financial institutions to apply for trial.

At this stage, how to develop the ability of big models in the financial sector, he believes that there are three challenges that need to be addressed:

The first challenge is that the ability of common models can not meet the needs of financial scenarios. First of all, the accuracy of the general large model itself is not enough, the second is the lack of financial knowledge of the general large model, and the third is the difficulty of updating and iteration of the general large model.

The second challenge is how to efficiently embed large models into existing business scenarios. A team should not only understand the know-how of business scenarios, but also understand the use of large models, but also need to have strong engineering capabilities, so that it is possible to select scenarios suitable for the application of large models and efficiently embed large models into the actual business process.

The third challenge is the security compliance and privacy issues arising from the application of large models in the financial industry. Finance itself is an industry with high compliance requirements, and the large model is a subversive new technology. How to balance the landing income of the large model and the potential compliance risk will be a more and more prominent problem.

How to solve these problems in the application of large models in the financial industry? Xu Dongliang believes that it is both unrealistic and uneconomical for each institution to solve these problems independently. It is very necessary for technology companies and financial institutions to cooperate on financial models. "Technology giants provide general models, or financial technology companies with strong technical capabilities provide large models of the financial industry." Specific financial institutions based on the underlying model of the industry, use their own business data to do private domain training, and then do the deployment and application of privatization.

In May this year, Lesser Fullness of Grain opened up the country's first 100 billion-level Chinese financial model, Xuanyuan, which has comprehensively surpassed the mainstream open source model in the financial domain task evaluation. Since open source, hundreds of financial institutions have applied for trial.

Lesser Fullness of Grain opened up the first 100 billion-level Chinese financial model "Xuanyuan" in China, by accelerating the landing of the big model, solving many difficult problems such as risk management and cost control by means of intelligent technology, and further improving the efficiency of financial services. Next, du Lesser Fullness of Grain will actively solve a series of problems in the application of large models in the financial industry, and contribute to the digital transformation of the financial industry.

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.