xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-09 06:05:24 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix: Correctly filter special tokens in benchmark_prefix_caching (#28615)

Browse Source 
Signed-off-by: Di Wu <dw2761@nyu.edu>
This commit is contained in:
Di Wu 2025-11-13 18:57:44 +08:00 committed by GitHub
parent 11ac9ddd03
commit e63fd44560
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
benchmarks/benchmark_prefix_caching.py
View File

@ -69,7 +69,7 @@ def sample_tokens(tokenizer: PreTrainedTokenizerBase, length: int) -> list[int]:
# Remove the special tokens.
return random.choices(
[v for k, v in vocab.items() if k not in all_special_ids],
[v for v in vocab.values() if v not in all_special_ids],
k=length,
)