Update app.py

app.py

@@ -1,19 +1,29 @@
 import gradio as gr
 from transformers import BlipProcessor, BlipForConditionalGeneration
+from gtts import gTTS
+import IPython.display as ipd
 
 model_id = "dblasko/blip-dalle3-img2prompt"
 model = BlipForConditionalGeneration.from_pretrained(model_id)
 processor = BlipProcessor.from_pretrained(model_id)
 
 def generate_caption(image):
-  inputs = processor(images=image, return_tensors="pt")
-  pixel_values = inputs.pixel_values
+    inputs = processor(images=image, return_tensors="pt")
+    pixel_values = inputs.pixel_values
 
-  generated_ids = model.generate(pixel_values=pixel_values, max_length=50)
-  generated_caption = processor.batch_decode(generated_ids, skip_special_tokens=True, temperature=0.8, top_k=40, top_p=0.9)[0]
+    generated_ids = model.generate(pixel_values=pixel_values, max_length=50)
+    generated_caption = processor.batch_decode(generated_ids, skip_special_tokens=True, temperature=0.8, top_k=40, top_p=0.9)[0]
 
-  return generated_caption
+    # Convert text to speech and save as audio file
+    tts = gTTS(text=generated_caption, lang='en')
+    tts.save("generated_audio.mp3")
 
-# Create a Gradio interface with an image input and a textbox output
-demo = gr.Interface(fn=generate_caption, inputs=gr.Image(), outputs=gr.Textbox(label="Generated caption"))
+    return generated_caption, "generated_audio.mp3"
+
+def play_audio(audio_path):
+    # Display an audio player
+    return ipd.Audio(audio_path)
+
+# Create a Gradio interface with an image input, a textbox output, and an audio player
+demo = gr.Interface(fn=generate_caption, inputs=gr.Image(), outputs=[gr.Textbox(label="Generated caption"), gr.Audio(player=True, label="Play Audio")])
 demo.launch()