#人工智能
#开源项目
前两天的Deepseek v3发布，效果有点震惊，看论文所说，只用了2048张H800显卡，2个月不到时间训练完毕，计算预算只有不到600万美元就训练出来了新的模型。
与其他大模型玩家不同的是，deepseek背后是一家叫幻方量化的公司，不是传统的互联网企业，前几个月看到的采访《揭秘DeepSeek:一个更极致的中国技术理想主义故事》，更早的一篇采访《疯狂的幻方：一家隐形AI巨头的大模型之路》。
BY codedump的电报频道
A project of our size needs at least a few hundred million dollars per year to keep going,” Mr. Durov wrote in his public channel on Telegram late last year. “While doing that, we will remain independent and stay true to our values, redefining how a tech company should operate.
For some time, Mr. Durov and a few dozen staffers had no fixed headquarters, but rather traveled the world, setting up shop in one city after another, he told the Journal in 2016. The company now has its operational base in Dubai, though it says it doesn’t keep servers there.Mr. Durov maintains a yearslong friendship from his VK days with actor and tech investor Jared Leto, with whom he shares an ascetic lifestyle that eschews meat and alcohol.
codedump的电报频道 from sg