Update app.py

app.py

@@ -6,6 +6,10 @@ feature_extractor = ViTFeatureExtractor.from_pretrained("nlpconnect/vit-gpt2-ima
 tokenizer = AutoTokenizer.from_pretrained("nlpconnect/vit-gpt2-image-captioning")
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+
+
+
+
 model.to(device)
 
 
@@ -34,5 +38,28 @@ def predict_step(image_paths):
 
 torch.hub.download_url_to_file('https://github.com/AaronCWacker/Yggdrasil/blob/main/images/35-Favorite-Games.jpg', '35-Favorite-Games.jpg')
 
-result = predict_step(['35-Favorite-Games.jpg'])
+#result = predict_step(['35-Favorite-Games.jpg'])
+
+def predict(image,max_length=64, num_beams=4):
+  image = image.convert('RGB')
+  image = feature_extractor(image, return_tensors="pt").pixel_values.to(device)
+  clean_text = lambda x: x.replace('<|endoftext|>','').split('\n')[0]
+  caption_ids = model.generate(image, max_length = max_length)[0]
+  caption_text = clean_text(tokenizer.decode(caption_ids))
+  return caption_text 
+
+description= "NLP Image Understanding"
+title = "NLP Image Understanding"
+article = "nlpconnect/vit-gpt2-image-captioning"
 
+interface = gr.Interface(
+        fn=predict,
+        inputs = input,
+        theme="grass",
+        outputs=output,
+        examples = ['35-Favorite-Games.jpg'],
+        title=title,
+        description=description,
+        article = article,
+    )
+interface.launch(debug=True)