This morning, Clement Delangue, co-founder and CEO of the AI model community Hugging Face, excitedly announced on social media that the DeepSeek-R1 model on their platform has received over 10,000 likes from users. This achievement has made it stand out among nearly 1.5 million models on Hugging Face, positioning it as one of the most popular large models. Delangue expressed his amazement by saying, "This result is truly astonishing!"

At the end of last month, Hugging Face also unveiled the Open-R1 project, aimed at further enhancing the open-source ecosystem of DeepSeek R1. The project plans to fill in the parts of DeepSeek R1 that have not yet been open-sourced, adding fuel to this globally popular model and driving its influence even further.

According to Hugging Face's official introduction, while the model weights used in DeepSeek are publicly available, the datasets and specific code used for training the model have remained unknown. Through the Open-R1 project, Hugging Face hopes to bridge these gaps and provide the community with more comprehensive resources, promoting deeper research and applications.

Previously reported, Hugging Face has highly praised DeepSeek R1, noting that it is not only efficient but can also serve as a foundational model for innovation, offering researchers, scientists, and enterprises an affordable choice to achieve breakthroughs and innovations in their respective fields.

In addition, Hugging Face revealed that the synthetic dataset they plan to establish will not be limited to mathematics alone but will extend to multiple other disciplines, providing strong support and assistance for science and other areas. This initiative undoubtedly will further enhance Hugging Face's influence and standing within the AI community.