From 1140991a7b3a435b06fb819b683cc12780a19e52 Mon Sep 17 00:00:00 2001 From: Russell Bryant Date: Fri, 14 Mar 2025 12:18:38 -0400 Subject: [PATCH] [V1] Fix vocab size calculation for structured output (#14826) Signed-off-by: Russell Bryant --- vllm/v1/structured_output/__init__.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vllm/v1/structured_output/__init__.py b/vllm/v1/structured_output/__init__.py index a341d74c5812b..32ea1852d0ac2 100644 --- a/vllm/v1/structured_output/__init__.py +++ b/vllm/v1/structured_output/__init__.py @@ -40,7 +40,7 @@ class StructuredOutputManager: tokenizer_group.ping() 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): # NOTE: ideally, xgrammar should handle this accordingly. # refer to https://github.com/mlc-ai/xgrammar/blob/d77c0a0173ef14779c918e3be7966ba852f7910f/python/xgrammar/tokenizer_info.py#L98