Error in creating audio data pipeline in Tensorflow

Question

I am trying to build an audio data pipeline in Tensor Flow. Below is the code.

def map_func(filename, label):
    sar=32000
    y, sr = librosa.load(filename,