Жители Санкт-Петербурга устроили «крысогон»17:52
FunctionGemma 是 Google 最小的函数调用专用模型——2.7 亿个参数，288 MB，解码速度约为 126 tok/s。没错，它需要微调（准确率从 58% 提升到 85%），没错，它使用了一种奇怪的自定义格式，而不是 JSON。但它适用于任何手机，响应速度极快，而且确实有效。现在就可以构建带有离线 AI 代理的应用——体积小、速度快、可靠性高，足以满足生产环境的需求。无需等待模型体积更小、设备速度更快的“神奇未来”，未来已来！
，推荐阅读51吃瓜获取更多信息
That Walmart deal was a smart way for Apple to test out the viability of cheaper MacBooks without building an entirely new product. But now the M1 Air’s design looks seriously dated, and the company also needs to move beyond the six-year-old M1 chip. It's time to get serious about delivering a true low-cost Apple laptop.
Mr Lemmens said that future spacecraft "should be designed in such a way that they can take themselves out of orbit safely, preferably by doing controlled re-entries".