ok just bad llama.cpp

vicuna-tools.d/main.go

@@ -322,7 +322,7 @@ func handleAPIChatBotPut(w http.ResponseWriter, r *http.Request) error {
 			defer Config.ChatBot.semaphore.Unlock()
 
 			if b, err := command.CombinedOutput(); err != nil {
-				return fmt.Errorf("error generating cache: %w: %s", err, b)
+				return fmt.Errorf("error generating cache with '%s': %w: %s", command.String(), err, b)
 			}
 			return nil
 		}(); err != nil {