llama.cpp
00381b07
- avoid copying the entire vector
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
1 year ago
avoid copying the entire vector
References
#6187 - llama_model_loader: support multiple split/shard GGUFs
#4 - Hp/split/load model (test CI)
Author
phymbert
Parents
1892ae7e
Loading