add TODO comment

Co-authored-by: Hyunjae Woo <[email protected]>

src/c++/perf_analyzer/genai-perf/genai_perf/llm_inputs/llm_inputs.py

@@ -640,6 +640,7 @@ def _get_input_dataset_from_file(cls, input_filename: Path) -> Dict:
         dataset_json["features"] = [{"name": "text_input"}]
         dataset_json["rows"] = []
         for prompt, image in zip(prompts, images):
+            # (TMA-2004) support variable images per request through input file
             content: Dict[str, Any] = {"text_input": prompt}
             content.update({"images": [image]} if image else {})
             dataset_json["rows"].append({"row": content})