vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-10 03:24:35 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

add metal mem limit

Browse source
This commit is contained in:
Lizonghang 2025-01-23 16:08:52 +04:00
parent 33429ec4e1
commit 78a544d716
5 changed files with 102 additions and 67 deletions
Download patch file Download diff file Expand all files Collapse all files

6
common/arg.cpp
View file

@ -731,10 +731,10 @@ gpt_params_context gpt_params_parser_init(gpt_params & params, llama_example ex,
}
).set_env("LLAMA_ARG_UNLOAD"));
add_opt(llama_arg(
{"-cm", "--cuda-mem"}, "N",
format("maximum cuda memory to use (default: %d)", params.cuda_mem),
{"-gm", "--gpu-mem"}, "N",
format("maximum GPU memory to use (default: %d)", params.gpu_mem),
[](gpt_params & params, int value) {
params.cuda_mem = value; // in GiB
params.gpu_mem = value; // in GiB
}
).set_env("LLAMA_ARG_CUDA_MEM"));
#ifdef GGML_USE_METAL