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 dce8e59ea..03915bf2d 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 ec019ad18..f9da65efa 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
@@ -280,8 +280,8 @@ class Data2VecAudioModel(BaseFairseqModel):
             torch.FloatTensor(cfg.encoder_embed_dim).uniform_()
         )
 
-        self.encoder = TransformerEncoder(cfg)
-        #self.encoder = TransformerEncoderAdapter(cfg)
+        #self.encoder = TransformerEncoder(cfg)
+        self.encoder = TransformerEncoderAdapter(cfg)
         self.layer_norm = LayerNorm(self.extractor_embed)
 
         self.final_proj = nn.Linear(self.embed, self.embed)