开启时会模糊预览图，关闭后正常显示
RT @kenanhsaleh: Excited to share that I’ve joined @a16z as an Investment Partner. I’ll be focusing on early-stage investing as part of a1…
GM @speedrun. Investor at @a16z. Apply here: https://t.co/bbxTvgMK6w You've probably played a game I worked on. Husband, girldad x2, cook, point guard
If you’re looking for an incredibly gifted, committed, senior full stack developer: Hire Lucas. Legit one of the best engineers I’ve ever worked with.
Product & Devs Growth @Cyfrin | Ex @Alchemy | Created @cyfrinupdraft and @AlchemyLearn | Robotics | Making web3 mainstream
RT @patrickc: Two conversations this weekend make me think that there's a vibe shift afoot in Silicon Valley around what one should work on…
GP @ a16z ... questionable heuristics in a grossly underdetermined world
最近在玩一个开源的 Voice Agent 框架，TEN Framework：https://t.co/TP0pNajYxu ，它有点像「实时语音 / 多模态 Agent 的操作系统」：在一套统一的实时流框架里，把 STT、LLM、TTS、VAD、Avatar 这些模块做成可插拔的“积木”，可以按需组合、替换，重点就是把和 AI 实时对话相关的低延迟、多模态和跨端部署这些工程问题打包解决掉。 花 10min 本地部署跑起来，还没有替换其他的东西，实际体验下它的实时语音问答挺丝滑：可以打断、响应够快，延迟大概在 1s 左右。像 Memory、RAG 这类常见能力也都已经帮你接好了，基于这些再往上可以扩展到 AI 情感陪伴、AI 口语陪练、电话 AI 客服、智能语音硬件这些场景。我用官方案例测试了一下，效果还不错，对「真·实时」语音 Agent 感兴趣的同学可以看看这个开源框架。
Father of Pake • MiaoYan • Mole • XRender
the handwritten note part may be a slight comedic exaggeration but i am amazed which papers are in the weights of claude enough to matter. insane pretraining data taste
i make things and do research https://t.co/jZh799yNH4 / https://t.co/IdaJwZJCXm
RT @rsdgpt: @antoine_chaffin how is colBERT this good? Subbed in my fine tune model and dropped RFF and my results got way better. You’ve s…
Asst professor @MIT EECS & CSAIL (@nlp_mit). Author of https://t.co/VgyLxl0oa1 and https://t.co/ZZaSzaRaZ7 (@DSPyOSS). Prev: CS PhD @StanfordNLP. Research @Databricks.