The Hong Kong Generative AI Research Center (HKGAI), funded by the Hong Kong Special Administrative Region government's "InnoHK Innovation Hong Kong R&D Platform," has officially launched its HKGAI V1 large language model. This is Hong Kong's first homegrown AI large language model, marking a new stage in the development of artificial intelligence in Hong Kong.
Image Source Note: Image generated by AI, image licensing provider Midjourney
HKGAI V1 is based on DeepSeek's full parameter fine-tuning technology and has undergone continuous training, successfully achieving the first Hong Kong localization of DeepSeek. While retaining the basic capabilities of DeepSeek V3 and R1, the model incorporates local Hong Kong knowledge and values, fully supporting Cantonese, English, and Mandarin Chinese, to better meet the diverse needs of the Hong Kong and overseas markets.
HKGAI has quickly built a complete DeepSeek ecosystem post-training technical system and completed localized training based on local Hong Kong data, forming Hong Kong's largest large model local knowledge base. It is understood that HKGAI V1 will be open-sourced soon, allowing Hong Kong citizens and developers to directly use this homegrown large language model product.
Currently, under the coordination of the Office of the Chief Technology Officer, approximately 70 departments of the Hong Kong Special Administrative Region government are using the large language model office assistance system provided by HKGAI, demonstrating its broad application potential in government scenarios. This launch not only injects new vitality into Hong Kong's AI ecosystem but also showcases the collaborative innovation capabilities between Hong Kong and mainland China in this field.