feat: clean output text

app.py

@@ -43,7 +43,7 @@ def inference_image_caption(raw_image):
     image = transform(raw_image).unsqueeze(0).to(device)   
     with torch.no_grad():
         caption = model.generate(image, sample=True, top_p=0.9, max_length=20, min_length=5)
-    return 'caption: '+caption[0]
+    return caption[0]
 
 inputs = gr.Image(type='pil')
 outputs = gr.outputs.Textbox(label="Output")