Update app.py

app.py

@@ -1,13 +1,17 @@
 import gradio as gr
 from transformers import pipeline
+from TTS.api import TTS
 
 captioner = pipeline(model="microsoft/git-base")
-tts = pipeline(model="suno/bark-small")
+tts = TTS(model_name="tts_models/multilingual/multi-dataset/your_tts", progress_bar=False, gpu=False)
 
 def predict(image):
     text = captioner(image)[0]["generated_text"]
-    audio = tts(text)["audio"]
-    return text, audio
+
+    audio_output = "output.wav"
+    tts.tts_to_file(text, speaker=tts.speakers[0], language="en", file_path=audio_output)
+    
+    return text, audio_output
 
 demo = gr.Interface(
     fn=predict,