Replies: 1 comment
-
I know there's a project called Petals where they do something like this, but I've never tried it. As far as I understand it, one of the big problems here is that a lot of data has to be moved around and the inference process (excluding prompt processing) is highly linear/sequential. So that doesn't lend itself very well to running distributed over the internet. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi, sorry in advance if this idea is really dumb (I'm a total noob in coding etc), but could it be possible, if a user allows it, to make a part of his GPU accessible to every other users when it is not used at times and if he has a good internet connexion? I was thinking that if a large number of users allows this, and with a good open source model, it could build a real free and decentralized GPT access for everybody.
Beta Was this translation helpful? Give feedback.
All reactions