Update inference.py

inference.py

@@ -30,10 +30,8 @@ def _calculate_f0(input: np.ndarray,length,sr,f0min,f0max,
 def get_text(wav,sr,transform=1.0):
 
     #wav, sr = librosa.load(file,sr=None)
-    if sr<16000:
-        return 'sample rate too low'
     if len(wav.shape) > 1:
-        wav = librosa.to_mono(wav) 
+        wav = librosa.to_mono(wav.transpose(1, 0)) 
     if sr!=16000:  
         wav16 = librosa.resample(wav, sr, 16000)
     else:
@@ -57,8 +55,10 @@ def getkey(key):
 def infer(f,speaker,key,reqf0=False):
     speaker=int(speaker[7:])
     if f is None:
-        return "You need to upload an audio", None
+        return "请上传音频", None
     sr, audio = f
+    if sr<16000:
+        return "采样率过低，请上传至少拥有16000Hz采样率的音频",None
     duration = audio.shape[0] / sr
     if duration > 120:
         return "请上传小于2min的音频", None