Update app.py

app.py

@@ -82,11 +82,11 @@ def tts_fn(slider_value, input_text):
         speaker_embedding = torch.tensor(speaker_embedding).unsqueeze(0)
 
     phonemes = PersianG2Pconverter.transliterate(input_text, tidy = False, secret = True)
-    # text = "</s>"
-    # for i in phonemes.replace(' .', '').split(" "):
-    #     text += i + " <pad> "
+    text = "</s>"
+    for i in phonemes.replace(' .', '').split(" "):
+        text += i + " <pad> "
 
-    text = phonemes
+    text += "</s>"
     
     print("sentence:", input_text)
     print("sentence phonemes:", text)
@@ -118,7 +118,8 @@ def tts_fn(slider_value, input_text):
 def master_fn(slider_value, input_text):
     if "." not in input_text:
         input_text += '.'
-        
+
+    print(f"speaker_id = {slider_value}")
     all_speech = []
     for sentence in input_text.split("."):
         sampling_rate_response, audio_chunk_response = tts_fn(slider_value, sentence)