Update app.py

app.py

@@ -1,10 +1,38 @@
 import gradio as gr
+from transformers import AutoImageProcessor, AutoTokenizer
+from optimum.onnxruntime import ORTModelForVision2Seq
+from PIL import Image
 
-with gr.Blocks(fill_height=True) as demo:
-    with gr.Sidebar():
-        gr.Markdown("# Inference Provider")
-        gr.Markdown("This Space showcases the WinKawaks/vit-small-patch16-224 model, served by the hf-inference API. Sign in with your Hugging Face account to use this API.")
-        button = gr.LoginButton("Sign in")
-    gr.load("models/WinKawaks/vit-small-patch16-224", accept_token=button, provider="hf-inference")
-    
-demo.launch()
+# Load model, tokenizer, and processor
+processor = AutoImageProcessor.from_pretrained("WinKawaks/vit-small-patch16-224")
+tokenizer = AutoTokenizer.from_pretrained("WinKawaks/vit-small-patch16-224")
+model = ORTModelForVision2Seq.from_pretrained(
+    "WinKawaks/vit-small-patch16-224", 
+    export=True
+)
+
+def run(image):
+    if image is None:
+        return "No image provided."
+
+    # Preprocess
+    inputs = processor(images=image, return_tensors="pt")
+
+    # Generate
+    outputs = model.generate(**inputs, max_length=64)
+
+    # Decode
+    text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return text
+
+
+# --- Gradio UI ---
+demo = gr.Interface(
+    fn=run,
+    inputs=gr.Image(type="pil"),
+    outputs="text",
+    title="ViT Vision2Seq ONNX Demo",
+    description="Upload an image → get generated text from WinKawaks/vit-small-patch16-224 (ONNX)."
+)
+
+demo.launch()