this post was submitted on 01 Dec 2024
46 points (88.3% liked)

Futurology

1823 readers
163 users here now

founded 1 year ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
46
Multiple LLMs voting together on content validation catch each other’s mistakes to achieve 95.6% accuracy. (arxiv.org)
submitted 3 days ago by [email protected] to c/[email protected]
27 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] pennomi 3 points 3 days ago (2 children)

A lot of the smaller LLMs don’t require GPU at all - they run just fine on a normal consumer CPU.

[–] [email protected] 1 points 1 day ago

yeah but 10x slower, at speeds that just don't work for many use cases. When you compare energy consumption per token, there isn't much difference.

[–] [email protected] 3 points 3 days ago (1 children)

Wouldn't running on a CPU (while possible) make it less energy efficient, though?

[–] pennomi 3 points 3 days ago

It depends. A lot of LLMs are memory-constrained. If you’re constantly thrashing the GPU memory it can be both slower and less efficient.