Recent submissions

Comment · 9 days ago
Imagine your printer has it's driver failing every 2 days but thankfully it can generate YouTube videos
Comment · 9 days ago
Typos are a sign of human intelligence :D
Comment · 9 days ago
this would be a nice challange but from what I did, very doable.. not a usable speed maybe.
Comment · 9 days ago
looks nice! what languages does it support?
Comment · 9 days ago
100% it's on my queue of devices
Comment · 9 days ago
happy to hear that, feel free to contribute.
Comment · 9 days ago
not usable ones tho, but something like Karapathy's LLM may run
Comment · 9 days ago
a sentient tamagotchi is not far from us..
Comment · 9 days ago
Limits pushed again this time on phones..

Qwen3.6-35B-A3B at 4.9-5.7 tok/s Qwen3-Next-80B-A3B at 2.8 tok/s

And targeting 10 tok/s for both, 35B modified is on hugging face and 80B is uploading..

Comment · 9 days ago
Neat, I like it, do you have github repo for it?
Comment · 9 days ago
I had this case in fine tuned small models (up to 27B) in finance and legal, over bigger models (cost-wise, on-prem for a regulated institution, and better perofrmance because we owned the data pipeline)
Comment · 9 days ago
Decent!! Also I love the idea of the live chat, would you mind if I put a reference to it on my repo?
Comment · 9 days ago
actually the pi zero is a nice target to play with next, will give a shot and report
Comment · 9 days ago
I'm also working on this thread in parallel because it has interesting business aspects, would be happy to see your progress when you release it
Comment · 9 days ago
I ran YOLO on an FPGA for my bachlor's project degree project like 7 years ago, but this is even more interesting, and I think we can push bigger models with weight streaming. I still have an FPGA lying around.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton