On May 10-11, the "EPAI Seed Program" was officially launched! The first "EPAI-based Large Model Application Practice Seminar" was successfully held in Beijing Zhigu Building. Inspur Information joined hands with company leaders, business directors and technical engineers of 20 Metabrain ecological partners to carry out practice and discussion exchanges around the application of large models. This seminar aims to provide better tool support and technical guidance for partners to develop large-scale model applications, promote the implementation of EPAI-based ecosystem construction and co-creation business models, and fully unleash the intelligent productivity of large-scale models.
EPAI (Enterprise Platform of AI) is an enterprise large model development platform built by Inspur Information, which can provide an efficient, easy-to-use and secure end-to-end development platform for the landing application of enterprise AI large models, which provides a series of tools such as data preparation, model training, knowledge retrieval, application framework, etc., supports the scheduling of multiple computing power and multi-model algorithms, and helps enterprises efficiently develop and deploy generative AI applications and create intelligent productivity.
Mao Bolin, general manager of the information channel promotion department of Inspur, said in the introduction of the EPAI ecosystem planning that this is a new co-creation business model, which is divided into three levels: promoting collaborative innovation and optimization of products on the product side, and creating scenario-based solutions; On the market side, promote the commercial implementation and replication of EPAI solutions in the industry; On the ecological side, relying on the regional meta-brain ecological alliance to promote more extensive ecological cooperation, everyone polymerizes with each other to form a chemical reaction, and jointly seize the valuable window period of the large model market.
Wu Shaohua, R&D Director of AI Software of Inspur Information, pointed out that when our partners use EPAI, we will invest in large model algorithm experts, software experts and other forces to fully empower partners, create benchmark applications, and promote the large-scale implementation of large model applications.
This seminar set up three parts: "Communication-Practice-Discussion" to share with you the practice of building a scenario-oriented large model application platform based on EPAI. Through explanations and practical operations, Inspur information technology experts let the participants learn how to use tools such as data cleaning, model fine-tuning, network search, knowledge base construction, and knowledge retrieval based on EPAI, and at the same time introduced the fine-tuning training practice of the industry's mainstream large models on the EPAI platform. At the end of the seminar, practical cases were shared, and the application of large models for enterprise scenarios was discussed.
Through two days of exchange and learning, the participants deepened their understanding of each module of EPAI, and through hands-on experience of tools such as network search, knowledge base construction and retrieval, the participants generally believed that EPAI can help solve the problems of complex enterprise large model application development process and high threshold. The partners hope to quickly cooperate with Inspur Information for POC, jointly create excellent large model application cases, and help large models expand a wider range of capabilities in enterprise application development and implementation through EPAI.
In the future, Inspur Information will further strengthen the EPAI ecosystem on the basis of the meta-brain ecology, help partners better integrate resources, improve R&D efficiency, optimize products and services, jointly promote the innovation and development of large model applications based on EPAI, and better release the commercial value of large models.