diff --git a/vllm/v1/worker/gpu_model_runner.py b/vllm/v1/worker/gpu_model_runner.py
index 3ea39dc519d8..bef67486d518 100644
--- a/vllm/v1/worker/gpu_model_runner.py
+++ b/vllm/v1/worker/gpu_model_runner.py
@@ -2192,7 +2192,7 @@ class GPUModelRunner(LoRAModelRunnerMixin, KVConnectorModelRunnerMixin):
                     high=self.model_config.get_vocab_size(),
                     dtype=input_ids.dtype)
 
-            logger.debug("Randomizing dummy data for DP Rank")
+            logger.debug_once("Randomizing dummy data for DP Rank")
             input_ids.copy_(rand_input_ids()[:input_ids.size(0)],
                             non_blocking=True)
             yield