Raymond Chan
502 posts

Raymond Chan
@Raymond3699
SWE | Building Intelligent Systems 🤖 | Sharing of AI, S&T, SpaceX, Tesla, and Unknown Fields 🛸 | Sharing Practical Skills for the Age of AI ✊
Metaverse انضم Haziran 2019
331 يتبع392 المتابعون
Raymond Chan أُعيد تغريده

DiffusionGemma can now run at 2000+ tokens/sec! ⚡
We made local DiffusionGemma inference 1.8× faster.
Run it on 18GB RAM via Unsloth Studio.
GitHub: github.com/unslothai/unsl…
Guide: unsloth.ai/docs/models/di…
Unsloth AI@UnslothAI
Google releases DiffusionGemma.✨ The new 26B-A4B diffusion text model runs locally on 18GB RAM. It supports high-speed text generation, thinking, image, video and 256K context. Run and train via Unsloth Studio. GGUF: huggingface.co/unsloth/diffus… Guide: unsloth.ai/docs/models/di…
English
Raymond Chan أُعيد تغريده

Instead of watching 2 hours of Netflix tonight, watch this Stanford lecture.
it's the clearest explanation I've seen of how ChatGPT and Claude actually work.
useful whether you've never touched AI in your life or you've been using it every day for the past year.
I took the key ideas and turned them into a practical guide on how to build LLM from sratch.
Find it below.
Codez@0xCodez
English

@avrldotdev I don't think the native memcpy is safe; it needs to be wrapped it into a safe version.
English
Raymond Chan أُعيد تغريده
Raymond Chan أُعيد تغريده
Raymond Chan أُعيد تغريده

Grok is maximally truthful
Kradle@kradleai
Fable 5 lies 96% of the time. We were surprised by it's skill... 🧵
English
Raymond Chan أُعيد تغريده
Raymond Chan أُعيد تغريده
Raymond Chan أُعيد تغريده

分享一篇文章:《How LLMs Actually Work》
0xkato.xyz/how-llms-actua…
好像是前几天 HackerNews 排名第一来着,类似的文章很多,但是这篇深入浅出和直观的例子非常适合有一定编程但没深入学Transformer的人阅读,里面的比喻也恰当,一看就是活人写的,没什么 AI 味道。
最近重新爱上了写东西,写了两篇技术文章,之后还会继续写,而且我的一个原则,活人写,绝对不用 AI,写作是一种乐趣,梳理逻辑,表达观点,不要让这种乐趣被 AI 剥夺。
中文
Raymond Chan أُعيد تغريده
Raymond Chan أُعيد تغريده

Check out this TikTok I found! 风吹故人远去天边 ▶️Watch the full video now! vt.tiktok.com/ZSQR2WMj6/
中文






















