From 1489cd6c02ac6b36b2014a89c5200f89086e5f49 Mon Sep 17 00:00:00 2001
From: Vincent Zhong <207368749+vincentzed@users.noreply.github.com>
Date: Sun, 14 Sep 2025 23:26:39 -0400
Subject: [PATCH] [docs / oneliner] update mmmu docs instruction (#9768)

Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
---
 benchmark/mmmu/README.md | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/benchmark/mmmu/README.md b/benchmark/mmmu/README.md
index 80db21921..61fea8bc4 100644
--- a/benchmark/mmmu/README.md
+++ b/benchmark/mmmu/README.md
@@ -39,8 +39,11 @@ You can use `--extra-request-body` to specify additional OpenAI request paramete
 python3 bench_sglang.py --extra-request-body '{"max_new_tokens": 128, "temperature": 0.01}'
 ```
 
-### Evaluate hf
+### Evaluate HF
 
 ```
 python benchmark/mmmu/bench_hf.py --model-path Qwen/Qwen2-VL-7B-Instruct
 ```
+
+# Profiling MMMU
+You should use the standard instructions found in the [dedicated profiling doc](../../docs/developer_guide/benchmark_and_profiling.md) if running this benchmark with the profile option. We recommend using `--concurrency 1` for consistency, which makes profiling and debugging easier.