[Frontend] Use new Renderer for Completions and Tokenize API (#32863)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
2026-01-31 20:51:15 +08:00
parent 8980001c93
commit f0a1c8453a
64 changed files with 2116 additions and 2003 deletions
--- a/vllm/model_executor/models/ultravox.py
+++ b/vllm/model_executor/models/ultravox.py
@@ -209,6 +209,7 @@ class UltravoxMultiModalProcessor(BaseMultiModalProcessor[UltravoxProcessingInfo
        item_processor_data = dict(**mm_data, audios=audios)

        # some tokenizer kwargs are incompatible with UltravoxProcessor
+        tok_kwargs.pop("add_special_tokens", None)
        tok_kwargs.pop("padding", None)
        tok_kwargs.pop("truncation", None)