biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc] Remove unused slot_mapping buffer (#23502)

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon
2025-08-24 14:03:36 -07:00
committed by GitHub
parent e2db1164a1
commit ad78868450
1 changed files with 0 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
vllm/v1/worker/gpu_model_runner.py
View File

@@ -254,9 +254,6 @@ class GPUModelRunner(LoRAModelRunnerMixin, KVConnectorModelRunnerMixin):
self.seq_lens = torch.zeros(self.max_num_reqs, self.seq_lens = torch.zeros(self.max_num_reqs,
dtype=torch.int32, dtype=torch.int32,
device=self.device) 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. # None in the first PP rank. The rest are set after load_model.
self.intermediate_tensors: Optional[IntermediateTensors] = None self.intermediate_tensors: Optional[IntermediateTensors] = None