vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 01:24:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

allow embeddings models to use mmap

Browse source
This commit is contained in:
Concedo 2025-06-07 10:14:00 +08:00
parent abc272d89f
commit cfcdfd69bd
3 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
otherarch/embeddings_adapter.cpp
View file

@ -117,7 +117,7 @@ bool embeddingstype_load_model(const embeddings_load_model_inputs inputs)
llama_model_params model_params = llama_model_default_params();
llama_context_params ctx_params = llama_context_default_params();
const int nthreads = inputs.threads;
model_params.use_mmap = false;
model_params.use_mmap = inputs.use_mmap;
model_params.use_mlock = false;
model_params.n_gpu_layers = inputs.gpulayers; //offload if possible
model_params.split_mode = llama_split_mode::LLAMA_SPLIT_MODE_LAYER;