ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-02-04 21:40:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

Be able to set a max. number of GPUs to be used in split mode graph (#1051)

Browse Source 
Co-authored-by: Iwan Kawrakow <iwan.kawrakow@gmail.com>
This commit is contained in:
Kawrakow
2025-12-11 07:22:53 +01:00
committed by GitHub
parent 6a5a707ac0
commit 9484d150d8
6 changed files with 67 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
src/llama.cpp
View File

@@ -1742,6 +1742,7 @@ static bool llm_load_tensors(
int mla_attn,
enum llama_split_mode split_mode,
int main_gpu,
int max_gpu,
const float * tensor_split,
bool use_mlock,
bool validate_quants,
@@ -1763,6 +1764,7 @@ static bool llm_load_tensors(
model.split_mode = split_mode;
model.main_gpu = main_gpu;
model.max_gpu = max_gpu;
model.n_gpu_layers = n_gpu_layers;
const int n_layer = hparams.n_layer;
@@ -2138,7 +2140,7 @@ static int llama_model_load(const std::string & fname, llama_model & model, llam
#endif
if (!llm_load_tensors(
ml, model, params.n_gpu_layers, params.mla, params.split_mode, params.main_gpu, params.tensor_split,
ml, model, params.n_gpu_layers, params.mla, params.split_mode, params.main_gpu, params.max_gpu, params.tensor_split,
params.use_mlock, params.validate_quants,
params.progress_callback, params.progress_callback_user_data
)) {
@@ -3985,6 +3987,7 @@ struct llama_model_params llama_model_default_params() {
/*.mla =*/ 0,
/*.split_mode =*/ LLAMA_SPLIT_MODE_LAYER,
/*.main_gpu =*/ 0,
/*.max_gpu =*/ 0,
/*.tensor_split =*/ nullptr,
/*.rpc_servers =*/ nullptr,
/*.progress_callback =*/ nullptr,