Upload app.py

app.py

@@ -23,7 +23,7 @@ translation_pipe = pipeline("translation", model="facebook/wmt19-en-ru")
 
 #rus text to rus speech transformation
 vits_model = VitsModel.from_pretrained("facebook/mms-tts-rus")
-vits_tokanizer = tokenizer = VitsTokenizer.from_pretrained("facebook/mms-tts-rus")
+vits_tokenizer = VitsTokenizer.from_pretrained("facebook/mms-tts-rus")
 
 def transform_audio_to_speech_en(audio):
     outputs = asr_pipe(audio, max_new_tokens=256, generate_kwargs={"task": "translate"})
@@ -35,7 +35,7 @@ def translator(text):
 
 def synthesise(translated_text):
     translated_text = translator(translated_text)
-    inputs = vits_tokanizer(translated_text, return_tensors="pt")
+    inputs = vits_tokenizer(translated_text, return_tensors="pt")
     with torch.no_grad():
         speech = vits_model(**inputs).waveform
     return speech.cpu()