use pyaudioconvert

nemo_asr.py

@@ -1,7 +1,7 @@
 import aiofiles
 import nemo
 import nemo.collections.asr as nemo_asr
-import librosa
+import pyaudioconvert as pac
 import soundfile as sf
 from pydub import AudioSegment
 
@@ -28,7 +28,6 @@ def transcribe(file, modelName="stt_rw_conformer_transducer_large"):
 #     print(transcription[0])
   sound = AudioSegment.from_file(file.name)
   sound.export("test.wav",format='wav')
-  x, sr = librosa.load('test.wav')
-  sf.write("Test2.wav", x, 16000)
+  pac.convert_wav_to_16bit_mono("Test2.wav","Test2.wav")
   transcription= asr_model.transcribe(["Test2.wav"])
   return  transcription[0]

requirements.txt

@@ -11,4 +11,5 @@ soundfile==0.10.3.post1
 wget
 aiofiles
 pydub
-git+https://github.com/NVIDIA/[email protected]#egg=nemo_toolkit[all]
+git+https://github.com/NVIDIA/[email protected]#egg=nemo_toolkit[all]
+pyaudioconvert