xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-08-01 11:54:27 +08:00
Code Issues Packages Projects Releases Wiki Activity

[V1] Fix vocab size calculation for structured output (#14826)

Browse Source 
Signed-off-by: Russell Bryant <rbryant@redhat.com>
This commit is contained in:
Russell Bryant 2025-03-14 12:18:38 -04:00 committed by GitHub
parent 613c5bb945
commit 1140991a7b
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
vllm/v1/structured_output/__init__.py
View File

@ -40,7 +40,7 @@ class StructuredOutputManager:
tokenizer_group.ping() tokenizer_group.ping()
tokenizer = tokenizer_group.get_lora_tokenizer(None) tokenizer = tokenizer_group.get_lora_tokenizer(None)
self.vocab_size = tokenizer.max_token_id self.vocab_size = tokenizer.max_token_id + 1
if isinstance(tokenizer, MistralTokenizer): if isinstance(tokenizer, MistralTokenizer):
# NOTE: ideally, xgrammar should handle this accordingly. # NOTE: ideally, xgrammar should handle this accordingly.
# refer to https://github.com/mlc-ai/xgrammar/blob/d77c0a0173ef14779c918e3be7966ba852f7910f/python/xgrammar/tokenizer_info.py#L98 # refer to https://github.com/mlc-ai/xgrammar/blob/d77c0a0173ef14779c918e3be7966ba852f7910f/python/xgrammar/tokenizer_info.py#L98