Update app.py

app.py

@@ -10,15 +10,12 @@ import os
 
 token = os.environ.get("HUGGINGFACE_HUB_TOKEN")
 
-
-
 model_id = "thelabel/240903-image-tagging"
-config = AutoConfig.from_pretrained(model_id,token=token)
-model = AutoModelForImageClassification.from_pretrained(model_id,token=token)
+config = AutoConfig.from_pretrained(model_id, token=token)
+model = AutoModelForImageClassification.from_pretrained(model_id, token=token)
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 model.to(device)
 
-# Standard ViT image transforms
 image_transform = transforms.Compose([
     transforms.Resize((224, 224)),
     transforms.ToTensor(),
@@ -30,7 +27,7 @@ def load_image_from_url(url):
         response = requests.get(url, timeout=10)
         response.raise_for_status()
         return Image.open(BytesIO(response.content)).convert("RGB")
-    except Exception as e:
+    except Exception:
         return None
 
 @spaces.GPU
@@ -55,9 +52,8 @@ def predict_tags(image_url, threshold=0.5):
 def gradio_predict(url, threshold):
     tags, error = predict_tags(url, threshold)
     if error:
-        return error, None
-    image = load_image_from_url(url)
-    return "\n".join([f"{tag}: {score:.2f}" for tag, score in tags]), image
+        return error
+    return "\n".join([f"{tag}: {score:.2f}" for tag, score in tags])
 
 demo = gr.Interface(
     fn=gradio_predict,
@@ -65,12 +61,10 @@ demo = gr.Interface(
         gr.Textbox(label="Image URL"),
         gr.Slider(0, 1, value=0.5, step=0.01, label="Threshold"),
     ],
-    outputs=[
-        gr.Textbox(label="Tags"),
-        gr.Image(label="Preview", type="pil"),
-    ],
+    outputs=gr.Textbox(label="Tags"),
     title="Image Tagging with ViT",
     description="Paste an image URL and get predicted tags using thelabel/240903-image-tagging model.",
 )
+
 if __name__ == "__main__":
     demo.launch()