llama.cpp
e9b7a5cb
- llama : use n_threads_batch only when n_tokens >= 32
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Hide Minimap (CTRL+M)
Commit
1 year ago
llama : use n_threads_batch only when n_tokens >= 32 ggml-ci
References
#4240 - llama : improve batched CPU perf with BLAS
Author
ggerganov
Committer
ggerganov
Parents
f815fe43
Files
1
llama.cpp
Loading