diff --git a/tools/who_what_benchmark/requirements.txt b/tools/who_what_benchmark/requirements.txt
index ba398cea50..c99f3a5ebb 100644
--- a/tools/who_what_benchmark/requirements.txt
+++ b/tools/who_what_benchmark/requirements.txt
@@ -7,7 +7,7 @@ pandas>=2.0.3
 numpy>=1.23.5
 tqdm>=4.66.1
 diffusers
-datasets==3.6.0
+datasets==4.1.1
 auto-gptq; sys_platform == "linux"
 autoawq<0.2.8; sys_platform == "linux"
 sentencepiece