asdasd

app.py

@@ -6,6 +6,7 @@ import torch
 model = VitsModel.from_pretrained("facebook/mms-tts-crh")
 tokenizer = AutoTokenizer.from_pretrained("facebook/mms-tts-crh")
 
+
 def predict(image):
     global model, tokenizer
     text = name
@@ -15,11 +16,11 @@ def predict(image):
     with torch.no_grad():
         output = model(**inputs).waveform
 
-    return output
+    return gr.Audio(output)
 
 gr.Interface(
     predict,
-    inputs= gr.Textbox(value="", label="Duration in seconds"),
+    inputs= gr.Textbox(value="", label="Duration in seconds"),,
     outputs="audio",
     title="Hot Dog? Or Not?",
 ).launch()