diff --git a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.data2vec_audio.py.swp b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.data2vec_audio.py.swp
index 090733575..f94443c1c 100644
Binary files a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.data2vec_audio.py.swp and b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/.data2vec_audio.py.swp differ
diff --git a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/data2vec_audio.py b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/data2vec_audio.py
index 7f0c5cf79..729fe6eef 100644
--- a/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/data2vec_audio.py
+++ b/egs/librispeech/ASR/pruned_transducer_stateless_d2v_v2/data2vec_audio.py
@@ -494,8 +494,10 @@ class Data2VecAudioModel(BaseFairseqModel):
             prompt = prompt.expand((features.size()[0], prompt.size()[0], prompt.size()[1]))
             features = torch.cat([prompt, features], dim=1)
             prompt_padding_mask = torch.zeros(prompt.size()[0], prompt.size()[1]).type(torch.BoolTensor).to(features.device)
-            padding_mask = torch.cat([prompt_padding_mask, padding_mask], dim=1)
-            print(padding_mask.size())
+            try: padding_mask = torch.cat([prompt_padding_mask, padding_mask], dim=1)
+            except:
+                print(prompt_padding_mask.size())
+                print(padding_mask.size())
 
         features = self.layer_norm(features)