Will's Blog: DeepSeek unveils DeepSeek-V3, a mixture-of-experts model of 671B total parameters, with 37B activated per token, claiming it outperforms top open-source models (Shubham Sharma/VentureBeat)

26 December 2024

DeepSeek unveils DeepSeek-V3, a mixture-of-experts model of 671B total parameters, with 37B activated per token, claiming it outperforms top open-source models (Shubham Sharma/VentureBeat)

Shubham Sharma / VentureBeat:
DeepSeek unveils DeepSeek-V3, a mixture-of-experts model of 671B total parameters, with 37B activated per token, claiming it outperforms top open-source models — Chinese AI startup DeepSeek, known for challenging leading AI vendors with its innovative open-source technologies, today released a new ultra-large model: DeepSeek-V3.

Posted from: this blog via Microsoft Power Automate.