China Unicom released a graphic model that can realize with text pictures, video clips / “LLM-AI” — phenomenal Lale moments

During the Shanghai Mobile World Communications Conference, China Unicom released the graphic model “Honghu Graphic Model 1.0”. China Unicom said that the Honghu graphic model is the first large-scale model for operators’ value-added services. It is understood that the Honghu graphic model currently has two versions, 800 million training parameters and 2 billion training parameters, which can realize functions such as text diagrams, video short films, and picture life maps. (YICAI report)

For large language models (LLM), they are currently capital-intensive competition, so the most common are the United States (ChatGPT) and China (ERNIE Bot). Become a leading group. At present, the most common thing in Taiwan is to directly use API string ChatGPT, applying ready-made General language models to become large-scale, otherwise they can only try on a small scale, but due to the high cost of computing power and storage requirements, many teams cannot survive to the day when they shine. So Lale APP is currently also trying to string ChatGPT, but users pay, please support us, thank you. (SDChain report)

SixDomainChain (SDChain) is the world’s first decentralized public blockchain ecosystem that integrates international standards of IoT Six-Domain Model and reference architecture standards for distributed blockchain.