diff --git a/README.md b/README.md
index 9908899..d65b530 100644
--- a/README.md
+++ b/README.md
@@ -15,7 +15,7 @@ The [default model weight](clmr_checkpoint_10000.pt) provided is pretrained on [
 
 ## Code Example
 
-Generate embeddings for the audio "test.wav". 
+Generate embeddings for the audio "test.wav".
 
 *Write the pipeline in simplified style*:
 
@@ -42,6 +42,7 @@ import towhee
           .audio_decode.ffmpeg['path', 'frames']()
           .runas_op['frames', 'frames'](func=lambda x:[y[0] for y in x])
           .audio_embedding.clmr['frames', 'vecs']()
+          .select['path', 'vecs']()
           .show()
 )
 ```
@@ -91,4 +92,4 @@ The input data should represent for an audio longer than 2s.
 *numpy.ndarray*
 
 Audio embeddings in shape (num_clips, 512).
-Each embedding stands for features of an audio clip with length of 2s.
\ No newline at end of file
+Each embedding stands for features of an audio clip with length of 2s.