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

fix

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon
2025-09-23 09:23:21 -07:00
parent 17c2c106b1
commit 42f99150c1
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/v1/worker/gpu/model_runner.py
View File

@ -293,7 +293,7 @@ class GPUModelRunner:
query_start_loc_gpu = query_start_loc.gpu[:num_reqs + 1]
query_start_loc_cpu = query_start_loc.cpu[:num_reqs + 1]
seq_lens_gpu = self.input_buffers.seq_lens.gpu[:num_reqs]
seq_lens_cpu = self.input_buffers.seq_lens.np[:num_reqs]
seq_lens_cpu = self.input_buffers.seq_lens.cpu[:num_reqs]
seq_lens_np = self.input_buffers.seq_lens.np[:num_reqs]
# Some input token ids are directly read from the last sampled tokens.