开启时会模糊预览图，关闭后正常显示
RT @StasBekman: Over the last year I have gathered lots of debug tech notes, so I'm expanding The Art of Debugging Open Book. Here is a ne…
RL and efficient distributed pretraining • eXperiments lab • memes and training lores
RT @horai93: Cloudflare Observability、InvocationをURLでshareできるようになった。最高。
building workers observability @cloudflaredev, prev founder @baselimehq (acquired by cloudflare), prev aerodynamicist
there is a mosquito in my apartment in november i am gonna crash out
codegen @mistralai, husband
RT @motiondotdev: Animating layout isn't always bad. Global CSS variables are a performance killer. Hardware accelerated blurs can crash si…
Creator of @tailwindcss. Listener of Slayer. Austin 3:16.
电脑里存放着大量文档、图片和 PDF，有时候想找某个资料都想不起来存放在哪个文件夹，很是苦恼。 不妨试试，利用 Second Brain 这个开源工具，把我们的本地文件变成一个可搜索的 RAG 知识库。 通过自然语言或关键词来搜索内容，支持文本和图片的多模态搜索，还能结果本地 AI 模型进行智能问答。 GitHub：https://t.co/7d2pQnINZo 主要特性： - 语义搜索和关键词搜索结合，精准找到相关内容； - 支持文本和图片的多模态搜索与嵌入； - 支持的文件类型，包括 txt、pdf、docx、png、gif 等多种格式； - 可选 AI 模式，支持本地模型和 OpenAI 云端模型； - 检索增强生成（RAG）让 AI 基于知识库给出高质量回答； - 完全本地运行保护隐私，支持同步 Google Drive。 支持使用 LM Studio 与本地部署的大模型结合使用，可处理超过 10 万个文件，所有数据均在本地处理。
💡 挖掘开源的价值 🧑🏻💻 坚持分享 GitHub 上高质量、有趣、实用的教程、AI工具、前沿 AI 技术 🧐 A list cool, interesting projects of GitHub. ✏️ 公众号：GitHubDaily
How Kimi-K2-Thinking compares to MiniMax M2 on size? 2/n 1. MiniMax M2 has 10B active and 230B total parameters with full attention. 2. Kimi K2 has 35B active and 1 trillion total parameters. Both have most of their weights in 8bits. That mean M2 will be much easier to host and its KV cache will be much more compact. MiniMax M2 uses full attention, would be interesting to see if Kimi-M2 has done something interesting to the attention layer. (for this calculations I am assuming Kimi-K2-Thinking is based on Kimi-K2-Base)
AI @amazon. All views personal!