Replying to
A 1551 loading a 108 block file with RAMBOard HypaRAM fastloader
https://www.youtube.com/watch?v=vWi2URbmIlo
Remote
ytm
@ytm@social.linux.pizza
mastodon 4.7.3Data science, math, ML, DIY and 8-bit computing
46 Followers
94 Following
4 Posts
Joined November 19, 2022
Replying to
@Megaboz Thanks. It's not quantized at all, these are all 32-bit floats. First layer run takes about 1.5 minutes with this model.
Featured on Hackaday:
https://hackaday.com/2025/05/03/llm-ported-to-the-c64-kinda/
"“If there’s one thing the Commodore 64 is missing, it’s a large language model,” is a phrase nobody has uttered on this Earth" That's my kind of humour.
#c64 #ai #gpt
