nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 23 Packages Projects Releases Wiki Activity

server: skip unused log lines on router mode (#24463)

Browse Source
This commit is contained in:
Xuan-Son Nguyen
2026-06-11 11:36:35 +02:00
committed by GitHub
parent 1bfbdb134e
commit 18ef86ecec
1 changed files with 14 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/server/server.cpp
+2
View File
@@ -94,6 +94,7 @@ int llama_server(int argc, char ** argv) {
const bool is_router_server = params.model.path.empty(); const bool is_router_server = params.model.path.empty();
common_params_print_info(params, !is_router_server); common_params_print_info(params, !is_router_server);
if (!is_router_server) {
// validate batch size for embeddings // validate batch size for embeddings
// embeddings require all tokens to be processed in a single ubatch // embeddings require all tokens to be processed in a single ubatch
// see https://github.com/ggml-org/llama.cpp/issues/12836 // see https://github.com/ggml-org/llama.cpp/issues/12836
@@ -109,6 +110,7 @@ int llama_server(int argc, char ** argv) {
params.n_parallel = 4; params.n_parallel = 4;
params.kv_unified = true; params.kv_unified = true;
} }
}
// for consistency between server router mode and single-model mode, we set the same model name as alias // for consistency between server router mode and single-model mode, we set the same model name as alias
if (params.model_alias.empty() && !params.model.name.empty()) { if (params.model_alias.empty() && !params.model.name.empty()) {