kanshan/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-06-26 14:20:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

server: fix remote preset handling, add test (#24938)

Browse Source 
* server: add test for remote preset

* fix remote preset handling

* fix

* fix test
This commit is contained in:
Xuan-Son Nguyen
2026-06-23 13:28:34 +02:00
committed by GitHub
parent c926ad0985
commit 75ad0b23ed
7 changed files with 52 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/server/server.cpp
+12 -1
View File
@@ -89,6 +89,17 @@ int llama_server(int argc, char ** argv) {
llama_backend_init();
llama_numa_init(params.numa);
// note: router mode also accepts -hf remote-preset, so we need to check that first
if (!params.model.hf_repo.empty()) {
try {
common_params_handle_models_params handle_params;
handle_params.preset_only = true;
common_params_handle_models(params, LLAMA_EXAMPLE_SERVER, handle_params);
} catch (const std::exception & e) {
// ignored for now
}
}
// router server never loads a model and must not touch the GPU
const bool is_router_server = params.model.path.empty()
&& params.model.hf_repo.empty();
@@ -263,7 +274,7 @@ int llama_server(int argc, char ** argv) {
return child.run_download(params);
} else if (!is_router_server) {
// single-model mode (NOT spawned by router)
common_params_handle_models(params, LLAMA_EXAMPLE_SERVER);
common_params_handle_models(params, LLAMA_EXAMPLE_SERVER, {});
}
//