Update utils.py

utils.py

@@ -7,6 +7,7 @@ from transformers import pipeline
 import re
 import torch
 import numpy as np
+import os
 
 _ref_audio_cache = {}
 asr_pipe = None
@@ -128,6 +129,12 @@ def preprocess_ref_audio_text(ref_audio_orig, ref_text, clip_short=True, show_in
 
     show_info("Converting audio...")
 
+    ref_audio_orig_converted = ref_audio_orig.replace(".wav", "_24k.wav").replace(".mp3", "_24k.mp3").replace(".m4a", "_24k.m4a").replace(".flac", "_24k.flac")
+
+    os.system("sox " + ref_audio_orig + " -r 24000 -c 1 " + ref_audio_orig_converted)
+
+    ref_audio_orig = ref_audio_orig_converted
+
     with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as f:
 
         aseg = AudioSegment.from_file(ref_audio_orig)