vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-07 20:49:03 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

disable device profiler in standalone mode

Browse source
This commit is contained in:
Lizonghang 2025-02-12 17:12:30 +04:00
parent c84f9d29fe
commit fdfaaecd5e
1 changed files with 4 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
common/common.cpp
View file

@ -1473,8 +1473,10 @@ struct llama_init_result llama_init_from_gpt_params(gpt_params & params) {
// get device profile // get device profile
LOG_INF("\nstart profiling this device, this may take some seconds ...\n"); LOG_INF("\nstart profiling this device, this may take some seconds ...\n");
dev_info.rank = params.rank; dev_info.rank = params.rank;
llama_profile_device(&dev_info, model, ml, params.gpu_mem, params.n_predict, params.n_ctx, params.cpuparams.n_threads, params.flash_attn); if (n_world > 1) {
llama_profile_device(&dev_info, model, ml, params.gpu_mem, params.n_predict, params.n_ctx, params.cpuparams.n_threads, params.flash_attn);
}
// create llama context // create llama context
struct llama_context_params cparams = llama_context_params_from_gpt_params(params); struct llama_context_params cparams = llama_context_params_from_gpt_params(params);
llama_context * lctx = llama_new_context_with_model(model, cparams); llama_context * lctx = llama_new_context_with_model(model, cparams);