From ad78868450a3596bed37dac05be9049019953e94 Mon Sep 17 00:00:00 2001 From: Woosuk Kwon Date: Sun, 24 Aug 2025 14:03:36 -0700 Subject: [PATCH] [Misc] Remove unused slot_mapping buffer (#23502) Signed-off-by: Woosuk Kwon --- vllm/v1/worker/gpu_model_runner.py | 3 --- 1 file changed, 3 deletions(-) diff --git a/vllm/v1/worker/gpu_model_runner.py b/vllm/v1/worker/gpu_model_runner.py index ed4a4e55f1212..ec9887b8010a2 100644 --- a/vllm/v1/worker/gpu_model_runner.py +++ b/vllm/v1/worker/gpu_model_runner.py @@ -254,9 +254,6 @@ class GPUModelRunner(LoRAModelRunnerMixin, KVConnectorModelRunnerMixin): self.seq_lens = torch.zeros(self.max_num_reqs, dtype=torch.int32, device=self.device) - self.slot_mapping = torch.zeros(self.max_num_tokens, - dtype=torch.int64, - device=self.device) # None in the first PP rank. The rest are set after load_model. self.intermediate_tensors: Optional[IntermediateTensors] = None