creates a FastAPI server and "mounts" our Gradio app inside

app.py

@@ -1,5 +1,6 @@
-# Final robust app.py for your Hugging Face Space
+# Final app.py using FastAPI wrapper
 
+from fastapi import FastAPI
 import gradio as gr
 import tensorflow as tf
 from huggingface_hub import hf_hub_download
@@ -8,7 +9,6 @@ from PIL import Image
 import os
 
 # --- 1. Load the Model ---
-model = None
 try:
     model_path = hf_hub_download(
         repo_id="skibi11/leukolook-eye-detector", 
@@ -18,43 +18,39 @@ try:
     print("--- MODEL LOADED SUCCESSFULLY! ---")
 except Exception as e:
     print(f"--- ERROR LOADING MODEL: {e} ---")
-    raise gr.Error(f"Failed to load model: {e}")
+    raise RuntimeError(f"Failed to load model: {e}")
 
-# --- 2. Pre-processing Logic ---
+# --- 2. Pre-processing & Prediction Logic (remains the same) ---
 def preprocess_image(img_pil):
     img = img_pil.resize((224, 224))
     img_array = np.array(img)
-    if img_array.ndim == 2:
-        img_array = np.stack((img_array,)*3, axis=-1)
-    if img_array.shape[-1] == 4:
-        img_array = img_array[..., :3]
+    if img_array.ndim == 2: img_array = np.stack((img_array,)*3, axis=-1)
+    if img_array.shape[-1] == 4: img_array = img_array[..., :3]
     img_array = img_array / 255.0
     img_array = np.expand_dims(img_array, axis=0)
     return img_array
 
-# --- 3. Prediction Logic ---
 def predict(image_from_gradio):
     if not isinstance(image_from_gradio, np.ndarray):
         return {"error": "Invalid input type. Expected an image."}
-
     try:
         pil_image = Image.fromarray(image_from_gradio)
         processed_image = preprocess_image(pil_image)
         prediction = model.predict(processed_image)
-
         labels = [f"Class_{i}" for i in range(prediction.shape[1])]
         confidences = {label: float(score) for label, score in zip(labels, prediction[0])}
         return confidences
     except Exception as e:
-        raise gr.Error(f"Error during prediction: {e}")
-
-# --- 4. Create Gradio Interface using gr.Blocks for stability ---
-with gr.Blocks() as demo:
-    gr.Interface(
-        fn=predict,
-        inputs=gr.Image(type="numpy"),
-        outputs=gr.JSON(),
-        api_name="predict"
-    )
-
-demo.launch()
+        return {"error": f"Error during prediction: {e}"}
+
+# --- 3. Create the Gradio Interface (without launching) ---
+gradio_interface = gr.Interface(
+    fn=predict,
+    inputs=gr.Image(type="numpy"),
+    outputs=gr.JSON(),
+    api_name="predict"
+)
+
+# --- 4. Create the FastAPI app and mount the Gradio app to it ---
+app = FastAPI()
+app = gr.mount_gradio_app(app, gradio_interface, path="/")