llama.cpp
c869d1e3
- CUDA: faster MMVQ for low-bit types
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
7 days ago
CUDA: faster MMVQ for low-bit types
References
#30517 - CUDA: faster MMVQ for low-bit types
Author
am17an
Parents
8c2134a8
Loading
