diff --git a/transformer_engine/pytorch/attention.py b/transformer_engine/pytorch/attention.py
index 5c2618b559..9268b9636e 100644
--- a/transformer_engine/pytorch/attention.py
+++ b/transformer_engine/pytorch/attention.py
@@ -604,7 +604,8 @@ def get_attention_backend(
             use_fused_attention = False
         elif cudnn_version >= (9, 6, 0) and qkv_format == "thd":
             logger.debug(
-                "Disabling FusedAttention as it does not support context parallelism with THD for cuDNN 9.6+"
+                "Disabling FusedAttention as it does not support context parallelism with THD for"
+                " cuDNN 9.6+"
             )
             use_fused_attention = False