xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-09 17:47:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon 2025-09-13 19:19:40 -07:00
parent e47bb9970b
commit eb3742c72a
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/v1/worker/gpu_block_table.py
View File

@ -118,7 +118,7 @@ class BlockTables:
# no clear upper bound on the number of new blocks. # no clear upper bound on the number of new blocks.
new_block_ids_cpu = torch.empty( new_block_ids_cpu = torch.empty(
self.num_kv_cache_groups, self.num_kv_cache_groups,
max(x[-1] for x in cu_num_new_blocks), max(len(x) for x in new_block_ids),
dtype=torch.int32, dtype=torch.int32,
device="cpu", device="cpu",
pin_memory=self.pin_memory, pin_memory=self.pin_memory,