change audio to 16Khz

nemo_asr.py

@@ -1,6 +1,7 @@
 import aiofiles
 import nemo
 import nemo.collections.asr as nemo_asr
+import librosa
 
 
 def transcribe(file, modelName="stt_rw_conformer_transducer_large"):
@@ -21,6 +22,8 @@ def transcribe(file, modelName="stt_rw_conformer_transducer_large"):
 #   for fname, transcription in zip(files, asr_model.transcribe(paths2audio_files=files)):
 #     print(f"Audio in {fname} was recognized as: {transcription}")
 #     print(transcription[0])
-  transcription= asr_model.transcribe([file.name])
+  x, sr = librosa.load(file.name)
+  librosa.output.write_wav("Test.wav", x, sr=16000, norm=False)
+  transcription= asr_model.transcribe(["Test.wav"])
   print(transcription)
   return {"text": transcription, "filename": file.name}

requirements.txt

@@ -7,7 +7,6 @@ stt
 webrtcvad
 numpy
 ffmpeg-python
-librosa==0.9.1
 soundfile==0.10.3.post1
 wget
 aiofiles