OPPO Xiaobu Assistant will have AI large model capabilities, based on AndesGPT

According to news from this site on August 13, following the first exposure of Huawei’s Hongmeng 4.0 to Xiaoyi’s AI large model and Xiaomi’s large model, OPPO and vivo have gradually begun to develop their efforts in the field of large models.

OPPO officially announced today that the new Xiaobu Assistant based on AndesGPT will soon launch a large-scale experience event.

It is said that the upgraded Xiaobu Assistant will have AI large model capabilities, stronger semantic understanding and dialogue capabilities, and can be customized according to needs. Copywriters can write content that users need, and AI capabilities such as summarizing and summarizing will also be greatly enhanced.

According to reports, AndesGPT is a generative large language model based on a hybrid cloud architecture created by the OPPO Andes Intelligent Cloud team. The team began to explore and implement pre-trained language models two years ago, and self-developed large models OBERT with 100 million, 300 million and 1 billion parameters. OBERT also once ranked among the Chinese language understanding evaluation benchmark CLUE1.1 Ranked fifth in the overall list and ranked first in the large-scale knowledge graph question and answer KgCLUE1.0 ranking.

OPPO revealed that AndesGPT has integrated key technologies such as continuous learning of Chinese dialogue data, fine-tuning of instructions, human feedback reinforcement learning and knowledge enhancement based on the OBERT model training experience. At the C-Eval in June this year It ranked fifth on the review list.

Inquiries on this site found that the OPPO Xiaobu Assistant team has already begun to work hard in the field of AI technology, including semantic understanding, dialogue generation, multi-modal and other generative AI-related technologies.

Zheng Zhitong, head of multi-modal learning at OPPO Xiaobu Intelligent Center, revealed on VALSE that OPPO has launched CV large-scale pre-training, NLP large-scale pre-training, multi-modal pre-training, and AIGC in the past two years. Drawing and AndesGPT and many other projects, and also demonstrated various business scenarios of OPPO Xiaobu Assistant on AIGC Drawing, including lock screen AI wallpaper, theme store AI wallpaper, Xiaobu space wallpaper in the App, cloud album search Beta, etc. Function, this shows how deeply OPPO is currently exploring the field of artificial intelligence.

In addition, OPPO also announced last month a free donation of the AI ​​project - ShaderNN to the LF AI&Data Foundation. According to reports, this is the industry’s first efficient and lightweight inference engine that supports the full graphics stack of OpenGL and Vulkan, targeting real-time AI applications for mobile graphics, images, and games.

Advertising statement: The external jump links (including but not limited to hyperlinks, QR codes, passwords, etc.) contained in the article are used to convey more information and save selection time. The results are for reference only. All articles on the site contain this statement.

The above is the detailed content of OPPO Xiaobu Assistant will have AI large model capabilities, based on AndesGPT. For more information, please follow other related articles on the PHP Chinese website!