←back to thread

578 points huseyinkeles | 2 comments | | HN request time: 0.47s | source
Show context
mhitza ◴[] No.45571218[source]
Should be "that you can train for $100"

Curios to try it someday on a set of specialized documents. Though as I understand the cost of running this is whatever GPU you can rent with 80GB of VRAM. Which kind of leaves hobbyists and students out. Unless some cloud is donating gpu compute capacity.

replies(2): >>45571268 #>>45571369 #
portaouflop ◴[] No.45571268[source]
If I have let’s say 40gb RAM does it not work at all or just take twice as long to train?
replies(1): >>45571442 #
1. typpilol ◴[] No.45571442[source]
Won't work at all. Or if it does it'll be so slow since it'll have to go to the disk for every single calculation so it won't ever finish.
replies(1): >>45572601 #
2. karpathy ◴[] No.45572601[source]
It will work great with 40GB GPU, probably a bit less than twice slower. These are micro models of a few B param at most and fit easily during both training and inference.