On October 30th, Facewall Intelligence and Baidu Intelligent Cloud announced a strategic partnership to jointly develop large-scale model cloud-edge collaboration solutions.
Under the collaboration agreement, both parties will leverage Baidu Intelligent Cloud's Qianfan Large Model Platform, Wenxin series of large models, and Facewall's MiniCPM series of edge-side large models, along with advanced technological accumulations in multiple terminals and various scenarios, to jointly develop and promote cloud-edge collaboration technology solutions. This initiative aims to enhance the performance of large models in practical applications such as embodied intelligence, smart terminals, and edge computing, reduce the inference costs of large models, improve model response speeds, and ultimately elevate user experience and business efficiency.
In September of this year, Facewall Intelligence released the new generation of edge-side large model—MiniCPM3.0, marking the arrival of the edge-side ChatGPT moment. With a model parameter size of 4 billion, MiniCPM3.0 has surpassed GPT-3.5 in natural language understanding, knowledge, code, and mathematics, and has distinguished itself among several well-known models both domestically and internationally.
In August, Facewall Intelligence also launched MiniCPM-V2.6, which reached a level fully comparable to GPT-4V, and for the first time enabled real-time video and multi-image joint understanding on the edge, garnering significant attention and accolades both domestically and internationally.