diff --git a/mmcv/ops/multi_scale_deform_attn.py b/mmcv/ops/multi_scale_deform_attn.py
index 546fb8b805b..7459263cdf0 100644
--- a/mmcv/ops/multi_scale_deform_attn.py
+++ b/mmcv/ops/multi_scale_deform_attn.py
@@ -178,7 +178,7 @@ class MultiScaleDeformableAttention(BaseModule):
     Args:
         embed_dims (int): The embedding dimension of Attention.
             Default: 256.
-        num_heads (int): Parallel attention heads. Default: 64.
+        num_heads (int): Parallel attention heads. Default: 8.
         num_levels (int): The number of feature map used in
             Attention. Default: 4.
         num_points (int): The number of sampling points for