alt Hacker Newsquantization has come a really long way. Q3 is fine, especially GSQ-RCO. You can run 27B well on 16GB of VRAM.
Replies
adornKey • today at 7:59 PM
With e.g. Strata you can run 125B Q4 very well on 16 GB of VRAM. 27B already has vintage vibes... OH, the good old 27B days...
The point of quantizing larger models is to beat other models running on the same limited-VRAM platform, i.e. 16GB VRAM in your example. You're doing "well" compared to other models meant for running on 16GB VRAM; you're not doing "well" at all compared to those running higher precisions for the same params up to at least Q8. (16 bit and higher is where the gain from further precision gets real iffy.)