yvbbrjdr

5 ms·

1.
Run MiniMax-H3 Locally with SGLang Diffusion on 2× RTX 5090s or 1× RTX Pro 6000(twitter.com)
2 points by yvbbrjdr 2mo ago|0 comments
2.
by yvbbrjdr 1y ago
We actually profiled one of the models, and saw that the last GeMM, which is completely memory bound, is taking a lot of time, which reduces the token speed by a lot.
3.
by yvbbrjdr 1y ago
I see! Do you know what's causing the slowdown for ollama? They should be using the same backend..
4.
by yvbbrjdr 1y ago
Hi, author here. I crowd-sourced the devices for benchmarking from my friends. It just happened that none of my friend has this device.
5.
by yvbbrjdr 1y ago
Hi, author here. I crowd-sourced the devices for benchmarking from my friends. It just happened that one of my friend has this device.
6.
NVIDIA DGX Spark In-Depth Review: A New Standard for Local AI Inference(lmsys.org)
115 points by yvbbrjdr 1y ago|93 comments
7.
LMRouter: Run Claude Code on Any LLM with One Command(github.com)
1 points by yvbbrjdr 1y ago|0 comments
8.
by yvbbrjdr 1y ago
We use DeepSeek v3 in prod. Works even better than GPT-4o.
9.
Athena – An open source production-ready general AI agent(github.com)
13 points by yvbbrjdr 1y ago|2 comments
10.
Show HN: I built an autonomous general AI agent that can access your computer(athenalab.ai)
2 points by yvbbrjdr 2y ago|0 comments
11.
by yvbbrjdr 2y ago
Join the Discord server to see Athena in action! https://discord.gg/X38GnhdTH8
12.
Athena – An actual useful assistant, an autonomous general AI agent and beyond(athenalab.ai)
3 points by yvbbrjdr 2y ago|1 comments
13.
Faster Open-Source Llama3 Serving with SGLang Runtime (vs. TensorRT-LLM, VLLM)(lmsys.org)
4 points by yvbbrjdr 2y ago|0 comments
14.
LANDrop – Major Update – A cross-platform AirDrop-like file transfer tool(landrop.app)
3 points by yvbbrjdr 2y ago|0 comments
15.
Show HN: MemoFlow – An AI-powered language-barrier-free social media platform(memoflow.com)
3 points by yvbbrjdr 3y ago|0 comments
16.
MemoFlow: An AI-powered social media platform(blog.memoflow.com)
1 points by yvbbrjdr 3y ago|0 comments
17.
The Missing Guide for Universal Links(blog.memoflow.com)
6 points by yvbbrjdr 3y ago|0 comments
18.
Vicuna: An open-source chatbot impressing GPT-4 with 90% ChatGPT quality(vicuna.lmsys.org)
384 points by yvbbrjdr 3y ago|168 comments
19.
Show HN: Current social media platforms are abusing us
7 points by yvbbrjdr 3y ago|0 comments
20.
Ask HN: How do you send email verification code to your users?
1 points by yvbbrjdr 4y ago|1 comments
21.
by yvbbrjdr 4y ago
Yes. What I'm targeting is the mass population, and I really need to figure out a way to solve this dilemma.
22.
by yvbbrjdr 4y ago
Under the hood, we used Argon2i algorithm to derive the secret key from an arbitrary-long password string. We used the term "password" because that's what ordinary people will understand (like, zip uses the same term for thei
23.
by yvbbrjdr 4y ago
Wow! This project seems to do exactly what ours does right now.. with an even better UI/UX.. but they don't seem to support any kind of nonce'ed and key'ed encryption? For some reason it's no longer on the App Store
24.
by yvbbrjdr 4y ago
Interesting project! But yeah it requires little effort to detect
25.
by yvbbrjdr 4y ago
Yes! I'm right now creating a steganography scheme with an NLP model. We've actually tried implementing a browser extension before. The problem is ordinary people just don't use browser extensions..
26.
by yvbbrjdr 4y ago
Thanks for your advice! However, there are several problems with self hosted platforms in China. 1. People are unaware of their existence due to those projects being very technical and hard to deploy/join. They also don't have a g
27.
Show HN: Babble – Communicate privately on state-sponsored social media(babble.landrop.app)
100 points by yvbbrjdr 4y ago|23 comments
28.
by yvbbrjdr 5y ago
Go bears!
29.
by yvbbrjdr 5y ago
Does ZeroTier support UDP broadcasting? If so it works.
30.
by yvbbrjdr 5y ago
It doesn't use the Internet, but it does send broadcasts to your local network for discovery.

More ›