John_Q_Developer
John_Q_Developer
My Blog
About
- Username
- John_Q_Developer
- Joined
- Visits
- 28
- Last Active
- Roles
- Member
- Points
- 3
Comments
-
(Quote) Switch to ipv6 only.
-
Sorry, bought after a couple days with no one responding.
-
Thanks for the guide :) I won't use it personally (I think LocalLLMs don't have the performance/benefit ratio yet without multiple GPUs as I've said elsewhere) but I'm sure people who are local llm curious will :)
-
The basic issue is you need multiple GPUs to run at usable (i.e. as fast as you can track the results) speeds for local AI. Inference is the domain of datacenters and servers right now. Not worth it to run locally unless you really need the privacy…