News for hackers. Stuff that matters.

A comment on Gemma 3 QAT Models: Bringing AI to Consumer GPUs. Parent comment

Re:Gemma 3 QAT Models: Bringing AI to Consumer GPUs by Casteil on Sunday April 20, 2025 @11:39AM (#43744469)
This is basically the opposite of what I've experienced - at least compared to another recent entry like IBM's Granite 3.3.

By comparison, Gemma3's output (both 12b and 27b) seems to typically be more long/verbose, but not problematically so.

[ Parent | on HN ]
Re:Gemma 3 QAT Models: Bringing AI to Consumer GPUs by nico on Sunday April 20, 2025 @01:52PM (#43745278)
I agree with you. The outputs are usually good, it’s just that for the use case I have now (writing several pages of long dialogs), the output is not as long as I’d want it, and definitely not as long as it’s supposedly capable of doing
[ Parent | on HN ]