vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-09 21:14:34 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

keep the output layer weights in shared memory by default

Browse source
This commit is contained in:
Lizonghang 2025-01-25 23:31:43 +04:00
parent f3dd5776eb
commit 1ca9a43bd1
4 changed files with 16 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

4
src/llama.cpp
View file

@ -19781,7 +19781,7 @@ struct llama_model_params llama_model_default_params() {
/*.use_mmap =*/ true,
/*.use_mlock =*/ false,
/*.check_tensors =*/ false,
/*.keep_out_in_metal =*/ false,
/*.keep_out_in_metal =*/ true,
};
#ifdef GGML_USE_METAL
@ -19799,7 +19799,7 @@ struct llama_context_params llama_context_default_params() {
/*.n_layer_window =*/ {32},
/*.n_gpu_layers =*/ 0,
/*.unload =*/ false,
/*.keep_out_in_metal =*/ false,
/*.keep_out_in_metal =*/ true,
/*.master_ip =*/ nullptr,
/*.next_node_ip =*/ nullptr,
/*.n_ctx =*/ 512,