Update app.py

app.py

@@ -22,12 +22,15 @@ def predict(image):
     image_tensor = feature_extractor(images=image, return_tensors="pt").pixel_values.to(device)
     caption_ids = caption_model.generate(image_tensor, max_length=128, num_beams=3)[0]
     caption_text = tokenizer.decode(caption_ids, skip_special_tokens=True)
-
+    return caption_text
+    
+def generate(image):
+    caption=predict(image)
     # Generate an image from the caption
-    generated_image = diffusion_model(caption_text)["sample"][0]
-
-    return caption_text, generated_images[0]
+    generated_image = diffusion_model(caption)["sample"][0]
 
+    return caption, generated_images[0]
+    
 # Set up Gradio interface
 input = gr.Image(label="Upload any Image", type='pil')
 outputs = [gr.Textbox(label="Caption"), gr.Image(label="Generated Image")]
@@ -41,4 +44,4 @@ interface = gr.Interface(
     #examples=examples,
     title=title,
 )
-interface.launch(debug=True)
+interface.launch(share=True)