Hacker News new | ask | show | jobs
Training a 67M-parameter transformer on an M4 Mac Mini (geddydukes.com)
3 points by Geddydukes 141 days ago
2 comments
recrush 141 days ago
Cool post. And you might find it interesting. https://github.com/KellerJordan/modded-nanogpt
link
Geddydukes 141 days ago
Thanks, I hadn't seen this repo before I'll check it out!
link
throwfaraway4 141 days ago
Cool post but the cursor shadow is a huge distraction
link
Geddydukes 141 days ago
Just went ahead and removed that, thanks for the comment and hopefully this will make it less distracting.
link