China’s Tencent has released a cloud service that supports constructing language models for every industry, equivalent to finance and education.

The South China Morning Post (SCMP) reported that Tencent announced a cloud service that gives a language model as a ‘MaaS (Model as Software)’ concept at a technology event held in Beijing on the nineteenth.

Tencent said it provides a one-stop solution within the cloud to assist customers pre-train and fine-tune language models and develop intelligent applications based on the models.

To this end, he explained that he opened a model store to gather high-quality language models specialized in various industries equivalent to finance, media, cultural tourism, government work and education. These models were created in collaboration with leading corporations in greater than 10 industries, including China’s state-owned CCTV, Fujian Big Data Group and Shanghai University, Tencent explained.

Customers can select an acceptable model amongst them, add their very own data, and fine-tune it to create a custom language model. In the method, they’re supported with a set of tools and high-performance computing clusters for data labeling, testing, evaluation, and deployment, in response to Tencent. Introduced.

Tencent has developed a big language model (LLM) called ‘Hunyian’ as a basic model equivalent to Open AI’s ‘ChatGPT’ and Alibaba’s ‘Hunybot’, but has not yet made it public. Based on this model, evidently they’ve released a service that gives a language model within the cloud.

As a cloud service that supports generative AI, including language models, Amazon Web Services (AWS) first launched Bedrock in April. It supports ‘Titan’, a big language model, ‘Clode’, an AI chatbot of Antropic, and ‘Stable Diffusion’, a picture creation tool.

Google Cloud also announced on the seventh that it would support LLM ‘Farm 2’, video creation tool ‘Imagine’, and code generation tool ‘Cody’ in ‘Vertex AI’, a machine learning platform.

Reporter Jeong Byeong-il jbi@aitimes.com