Lumpiasty/llama.cpp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

llama: add default load-mode auto, which avoids mmap on iGPUs (#26081)

Browse Source 
* llama: add new default load-mode auto which picks mmap unless a non-Metal iGPU is used

* Update ggml/src/ggml-hexagon/ggml-hexagon.cpp

Co-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com>

* set mmap_support to false on OpenCL backend

* fix order of load modes

* use -1 for auto

* resolve load mode auto earlier to correctly pick gpu host or cpu memory

* add load mode auto to llama-bench

* bump virtgpu api version, regenerate docs

---------

Co-authored-by: Piotr Wilkin (ilintar) <piotr.wilkin@syndatis.com>
Co-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Ruben Ortlam
2026-08-11 09:20:46 +03:00
committed by GitHub
co-authored by Max Krasnyansky Piotr Wilkin Georgi Gerganov
parent b3df57286c
commit 153d324bcf
33 changed files with 75 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files
common/arg.cpp
+4 -2
View File
@@ -2605,14 +2605,16 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
).set_env("LLAMA_ARG_DIO"));
add_opt(common_arg(
{"-lm", "--load-mode"}, "MODE",
"model loading mode (default: mmap)\n"
"model loading mode (default: auto)\n"
"- auto: mmap, unless a device does not support it\n"
"- none: no special loading mode\n"
"- mmap: memory-map model (if mmap disabled, slower load but may reduce pageouts if not using mlock)\n"
"- mlock: force system to keep model in RAM rather than swapping or compressing\n"
"- mmap+mlock: mmap + force system to keep model in RAM rather than swapping or compressing\n"
"- dio: use DirectIO if available\n",
[](common_params & params, const std::string & value) {
/**/ if (value == "none") { params.load_mode = LLAMA_LOAD_MODE_NONE; }
/**/ if (value == "auto") { params.load_mode = LLAMA_LOAD_MODE_AUTO; }
else if (value == "none") { params.load_mode = LLAMA_LOAD_MODE_NONE; }
else if (value == "mmap") { params.load_mode = LLAMA_LOAD_MODE_MMAP; }
else if (value == "mlock") { params.load_mode = LLAMA_LOAD_MODE_MLOCK; }
else if (value == "mmap+mlock") { params.load_mode = LLAMA_LOAD_MODE_MMAP_MLOCK; }
common/common.h
+1 -1
View File
@@ -473,7 +473,7 @@ struct common_params {
std::vector<size_t> fit_params_target = std::vector<size_t>(llama_max_devices(), 1024 * 1024*1024);
enum llama_split_mode split_mode = LLAMA_SPLIT_MODE_LAYER; // how to split the model across GPUs
enum llama_load_mode load_mode = LLAMA_LOAD_MODE_MMAP; // how to load the model
enum llama_load_mode load_mode = LLAMA_LOAD_MODE_AUTO; // how to load the model
common_cpu_params cpuparams;
common_cpu_params cpuparams_batch;