Skip to content

LLM inference in C/C++

C++ ggml

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

Xuan-Son Nguyen18ef86ecec72
diff --git a/tools/server/server.cpp b/tools/server/server.cpp
index a6ea749d0c..da635c6256 100644
--- a/tools/server/server.cpp
+++ b/tools/server/server.cpp
@@ -94,20 +94,22 @@ int llama_server(int argc, char ** argv) {
const bool is_router_server = params.model.path.empty();
common_params_print_info(params, !is_router_server);
- // validate batch size for embeddings
- // embeddings require all tokens to be processed in a single ubatch
- // see https://github.com/ggml-org/llama.cpp/issues/12836
- if (params.embedding && params.n_batch > params.n_ubatch) {
- SRV_WRN("embeddings enabled with n_batch (%d) > n_ubatch (%d)\n", params.n_batch, params.n_ubatch);
- SRV_WRN("setting n_batch = n_ubatch = %d to avoid assertion failure\n", params.n_ubatch);
- params.n_batch = params.n_ubatch;
- }
+ if (!is_router_server) {
+ // validate batch size for embeddings
+ // embeddings require all tokens to be processed in a single ubatch
+ // see https://github.com/ggml-org/llama.cpp/issues/12836
+ if (params.embedding && params.n_batch > params.n_ubatch) {
+ SRV_WRN("embeddings enabled with n_batch (%d) > n_ubatch (%d)\n", params.n_batch, params.n_ubatch);
+ SRV_WRN("setting n_batch = n_ubatch = %d to avoid assertion failure\n", params.n_ubatch);
+ params.n_batch = params.n_ubatch;
+ }
- if (params.n_parallel < 0) {
- SRV_INF("%s", "n_parallel is set to auto, using n_parallel = 4 and kv_unified = true\n");
+ if (params.n_parallel < 0) {
+ SRV_INF("%s", "n_parallel is set to auto, using n_parallel = 4 and kv_unified = true\n");
- params.n_parallel = 4;
- params.kv_unified = true;
+ params.n_parallel = 4;
+ params.kv_unified = true;
+ }
}
// for consistency between server router mode and single-model mode, we set the same model name as alias