[VLM] Remove BaseProcessingInfo.get_mm_max_tokens_per_item (#16408)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/model_executor/models/qwen_vl.py

@@ -530,13 +530,6 @@ class QwenVLProcessingInfo(BaseProcessingInfo):
     def get_supported_mm_limits(self) -> Mapping[str, Optional[int]]:
         return {"image": None}
 
-    def get_mm_max_tokens_per_item(
-        self,
-        seq_len: int,
-        mm_counts: Mapping[str, int],
-    ) -> Mapping[str, int]:
-        return {"image": self.get_num_image_tokens()}
-
     def get_num_image_tokens(self) -> int:
         hf_config = self.get_hf_config()
         vision_config = hf_config.visual