Update app.py

app.py

@@ -29,8 +29,10 @@ def text2speech(model, text):
         speaker_embedding = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)
         
         speech = synthesiser(text, forward_params={"speaker_embeddings": speaker_embedding})
-        print(speech, type(speech) )
-        return speech["audio"], speech["sampling_rate"]
+        audio_data = np.frombuffer(speech["audio"], dtype=np.float32)
+        audio_data_16bit = (audio_data * 32767).astype(np.int16)
+    
+        return sampling_rate, audio_data_16bit
 
 radio1 = gr.Radio(["microsoft/resnet-50", "google/vit-base-patch16-224", "apple/mobilevit-small"], label="Select a Classifier", info="Image Classifier")
 tab1 = gr.Interface(
@@ -50,7 +52,7 @@ radio3 = gr.Radio(["microsoft/speecht5_tts"], label="Select an tts", info="Age C
 tab3 = gr.Interface(
     fn=text2speech,
     inputs=[radio3, "text"],
-    outputs=[gr.Audio(label="Generated Speech", type="numpy")],
+    outputs=["audio"],
 )
 
 demo = gr.TabbedInterface([tab1, tab2, tab3], ["tab1", "tab2", "tab3"])