A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 23:15 UTC
[Misc] Remove useless code in cpu_worker (#5824)
Jie Fu (傅杰) committed
Jun 26, 2024 at 00:41 UTC
7b993143014c95844b380a5b05eebd14ad77b7aa
1 file changed
-1
vllm/worker/cpu_worker.py
-1
@@ -277,7 +277,6 @@ class CPUWorker(LoraNotSupportedWorkerBase):
277
assert seq_group_metadata_list is not None
278
num_seq_groups: int = len(seq_group_metadata_list)
279
assert execute_model_req is not None
280
- blocks_to_copy = execute_model_req.blocks_to_copy
280
blocks_to_copy = torch.tensor(execute_model_req.blocks_to_copy,
281
device="cpu",
282
dtype=torch.int64).view(-1, 2)