llama: only use one iGPU device by default (#23897)

diff --git a/src/llama.cpp b/src/llama.cpp
index edacd1d5f424ee0a11163f4f51474d39c9d1c936..a67fa8039a48df3415bee04d4c0bae9c2db02906 100644 (file)
--- a/src/llama.cpp
+++ b/src/llama.cpp
@@ -225,7 +225,9 @@ static bool llama_prepare_model_devices(const llama_model_params & params, llama
                     }
 
                     case GGML_BACKEND_DEVICE_TYPE_IGPU:
-                        igpus.push_back({false, dev});
+                        if (igpus.empty()) {
+                            igpus.push_back({false, dev});
+                        }
                         break;
                     case GGML_BACKEND_DEVICE_TYPE_META:
                         GGML_ABORT("fatal error");