Update app.py

app.py

@@ -1,37 +1,31 @@
-import os
-# Force legacy Keras logic
-os.environ["TF_USE_LEGACY_KERAS"] = "1"
-
 import gradio as gr
 import numpy as np
 import tensorflow as tf
-import tf_keras  # This is the dedicated legacy library
+from PIL import Image
 
-# Load the model using the legacy-specific loader
-model = tf_keras.models.load_model("model.h5", compile=False)
+# Standard loading - TF 2.12 will handle your .h5 file correctly
+model = tf.keras.models.load_model("model.h5", compile=False)
 
-# Labels for your waste classification
 labels = ["Cardboard", "Glass", "Metal", "Paper", "Plastic", "Trash"]
 
 def classify_waste(image):
-    # Preprocessing: Match the 224x224 input size from your error log
+    # Resize and normalize
     img = image.resize((224, 224))
     img_array = np.array(img) / 255.0
     img_array = np.expand_dims(img_array, axis=0)
     
-    # Run prediction
+    # Predict
     predictions = model.predict(img_array)[0]
     
-    # Return top results in Gradio format
+    # Create dictionary for Gradio
     return {labels[i]: float(predictions[i]) for i in range(len(labels))}
 
-# Setup the interface
 demo = gr.Interface(
     fn=classify_waste,
     inputs=gr.Image(type="pil"),
     outputs=gr.Label(num_top_classes=3),
     title="♻️ AI Waste Classifier",
-    description="Upload an image of waste to see the classification."
+    description="Upload an image to identify waste categories."
 )
 
 demo.launch()