[Bugfix] Fix Token IDs Reference for MiniCPM-V When Images are Provided With No Placeholders (vllm-project#8991)

Signed-off-by: Alex-Brooks <[email protected]>
Signed-off-by: Alvant <[email protected]>

Author: alex-jw-brooks

vllm/model_executor/models/minicpmv.py

@@ -274,8 +274,8 @@ def get_placeholder(image_size: Tuple[int, int], num_image: int):
             get_slice_image_placeholder(image_size, num_image)
 
     prompt = llm_inputs.get("prompt")
+    token_ids = llm_inputs.get("prompt_token_ids")
     if prompt is None:
-        token_ids = llm_inputs.get("prompt_token_ids")
         prompt = tokenizer.decode(token_ids)
 
     pattern = "(<image>./</image>)"