llama.cpp
33a004e9
- llama : more metal-friendly KV cache PAD
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
1 year ago
llama : more metal-friendly KV cache PAD
References
mlx-challenge
Author
ggerganov
Committer
ggerganov
Parents
b1f8af18
Loading