[V1] log GPU blocks num for MultiprocExecutor (vllm-project#11656)

neuralmagic · Jan 4, 2025 · e5d7ed0 · e5d7ed0
1 parent ad0d567
commit e5d7ed0
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/vllm/v1/executor/multiproc_executor.py b/vllm/v1/executor/multiproc_executor.py
@@ -95,6 +95,7 @@ def initialize(self, num_gpu_blocks: int) -> None:
         Initialize the KV caches and begin the model execution loop of the
         underlying workers.
         """
+        logger.info("# GPU blocks: %d", num_gpu_blocks)
         self.collective_rpc("initialize_cache", args=(num_gpu_blocks, ))
         self.collective_rpc("compile_or_warm_up_model")