Update app.py

app.py

@@ -1,5 +1,5 @@
 import gradio as gr
-from transformers import AutoProcessor, AutoModelForCausalLM, BlipForConditionalGeneration
+from transformers import AutoProcessor, AutoImageProcessor, AutoModelForCausalLM, BlipForConditionalGeneration
 import torch
 
 torch.hub.download_url_to_file('http://images.cocodataset.org/val2017/000000039769.jpg', 'cats.jpg')
@@ -11,6 +11,9 @@ git_model = AutoModelForCausalLM.from_pretrained("microsoft/git-base-coco")
 blip_processor = AutoProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
 blip_model = BlipForConditionalGeneration.from_pretrained("Salesforce/blip-image-captioning-base")
 
+vitgpt_processor = AutoImageProcessor.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+vitgpt_model = VisionEncoderDecoderModel.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
+
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
 git_model.to(device)
@@ -31,7 +34,9 @@ def generate_captions(image):
 
     caption_blip = generate_caption(blip_processor, blip_model, image)
 
-    return caption_git, caption_blip
+    caption_vitgpt = generate_caption(vitgpt_processor, vitgpt_model, image)
+
+    return caption_git, caption_blip, caption_vitgpt
 
    
 examples = [["cats.jpg"], ["stop_sign.png"]]
@@ -42,7 +47,7 @@ article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2102.033
 
 interface = gr.Interface(fn=generate_captions, 
                          inputs=gr.inputs.Image(type="pil"),
-                         outputs=[gr.outputs.Textbox(label="Caption generated by GIT"), gr.outputs.Textbox(label="Caption generated by BLIP")], 
+                         outputs=[gr.outputs.Textbox(label="Caption generated by GIT"), gr.outputs.Textbox(label="Caption generated by BLIP"), gr.outputs.Textbox(label="Caption generated by ViT+GPT-2")], 
                          examples=examples, 
                          title=title,
                          description=description,