[Core] [Bugfix] [Multimodal] Fix multimodal profiling and generation for SFT/PTQed models (#20058)

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

vllm/model_executor/models/qwen_vl.py

@@ -580,6 +580,7 @@ class QwenVLMultiModalProcessor(BaseMultiModalProcessor[QwenVLProcessingInfo]):
         prompt: str,
         mm_data: Mapping[str, object],
         mm_kwargs: Mapping[str, object],
+        tok_kwargs: Mapping[str, object],
     ) -> BatchFeature:
         # Drops anything between <img>/</img> tags; encoding with the tokenizer
         # will automatically add the image pads for the context.
@@ -600,6 +601,7 @@ class QwenVLMultiModalProcessor(BaseMultiModalProcessor[QwenVLProcessingInfo]):
             prompt=prompt,
             mm_data=mm_data,
             mm_kwargs=mm_kwargs,
+            tok_kwargs=tok_kwargs,
         )
 
     def _hf_processor_applies_updates(
@@ -607,6 +609,7 @@ class QwenVLMultiModalProcessor(BaseMultiModalProcessor[QwenVLProcessingInfo]):
         prompt_text: str,
         mm_items: MultiModalDataItems,
         hf_processor_mm_kwargs: Mapping[str, object],
+        tokenization_kwargs: Mapping[str, object],
     ) -> bool:
         return False