Update app.py

app.py

@@ -32,9 +32,13 @@ def speech_to_text(input_audio_or_text):
     return transcription.strip()
 
 def text_to_speech(text):
+    # Ensure the text input is not empty to avoid padding errors in the transformer model
+    if not text.strip():
+        return "The text input is empty, please provide a valid string."
+
     text = text.lower().translate(str.maketrans('', '', string.punctuation))
     inputs = tts_tokenizer(text, return_tensors="pt")
-    inputs['input_ids'] = inputs['input_ids'].long()  # Corrected assignment for input_ids
+    inputs['input_ids'] = inputs['input_ids'].long()  # Ensure input_ids are of type Long
     with torch.no_grad():
         output = tts_model(**inputs).waveform
     waveform = output.numpy().squeeze()
@@ -59,4 +63,4 @@ iface = gr.Interface(
     description="This app converts speech to text and then back to speech, ensuring the output audio is resampled to 16kHz."
 )
 
-iface.launch(share=True)  # Added `share=True` for creating a public link
+iface.launch(share=True)