llama.cpp
92472ea2
- cuda : unroll some of the loops
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
1 year ago
cuda : unroll some of the loops
References
#5021 - ggml : add Flash Attention
Author
ggerganov
Parents
1f8a5924
Loading
