vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-10 10:14:33 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

update README

Browse source
This commit is contained in:
Li, Zonghang 2025-06-16 12:09:07 +04:00
parent 45de284f3d
commit dfb1feb54e
3 changed files with 28 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

5
examples/speculative/speculative.cpp
View file

@ -70,8 +70,9 @@ int main(int argc, char ** argv) {
gpt_params params_draft = params;
params_draft.model = params_draft.model_draft;
params_draft.n_gpu_layers = params_draft.n_gpu_layers_draft;
params_draft.n_world = 1; // do not split the draft model across devices
params_draft.rank = 0; // always load the draft model on the head device
params_draft.n_world = 1; // do not split the draft model across devices
params_draft.rank = 0; // always load the draft model on the head device
params_draft.use_mlock = true; // always use mlock for the draft model
std::fill_n(params_draft.n_layer_window, params.n_world, 0);
if (params_draft.draft_cpuparams.n_threads > 0) {