Create app.py

app.py

@@ -0,0 +1,36 @@
+import gradio as gr
+import tensorflow as tf
+import numpy as np
+
+# Load the model
+model = tf.keras.models.load_model('stomach.h5')
+
+# Define the class names
+class_names = {
+    0: 'Esophagitis',
+    1: 'Dyed lifted polyps'
+}
+
+
+def classify_image(image):
+    # Preprocess the image
+    img_array = tf.image.resize(image, [256, 256])
+    img_array = tf.expand_dims(img_array, 0) / 255.0
+
+    # Make a prediction
+    prediction = model.predict(img_array)
+    predicted_class = tf.argmax(prediction[0], axis=-1)
+    confidence = np.max(prediction[0])
+
+    return class_names[predicted_class.numpy()], confidence
+
+
+iface = gr.Interface(
+    fn=classify_image,
+    inputs="image",
+    outputs=["text", "number"],
+    examples=[
+        ['examples/0.jpg'],
+        ['examples/1.jpg'],
+    ])
+iface.launch()