Create app.py

app.py

@@ -0,0 +1,50 @@
+import streamlit as st
+import numpy as np
+from keras.preprocessing import image
+from keras.models import load_model
+
+# Define a dictionary of classes
+classes = {'french_bulldog': 0,
+           'german_shepherd': 1,
+           'golden_retriever': 2,
+           'poodle': 3,
+           'yorkshire_terrier': 4}
+
+# Load the saved model from the disk
+model = load_model('best_model.h5')
+
+# Define the function for predicting the dog breed
+def predict_dog_breed(image_path):
+    # Load the image from the specified path and preprocess it
+    img = image.load_img(image_path, target_size=(256, 256))
+    x = image.img_to_array(img)
+    x = np.expand_dims(x, axis=0)
+    x = x / 255.
+
+    # Make the prediction using the trained model
+    preds = model.predict(x)
+    class_idx = np.argmax(preds[0])
+    predicted_class = [k for k, v in classes.items() if v == class_idx][0]
+
+    # Return the predicted dog breed
+    return predicted_class
+
+# Define the Streamlit app
+def app():
+    st.title("Dog Breed Classification App")
+    st.write("Upload an image of a dog to predict its breed.")
+
+    # Allow the user to upload an image file
+    uploaded_file = st.file_uploader("Choose a dog image...", type=["jpg", "jpeg", "png"])
+
+    # Make the prediction when the user clicks the "Predict" button
+    if uploaded_file is not None:
+        image_path = f"tmp/{uploaded_file.name}"
+        with open(image_path, "wb") as f:
+            f.write(uploaded_file.getbuffer())
+        predicted_class = predict_dog_breed(image_path)
+        st.write(f"Predicted dog breed: {predicted_class}")
+
+# Run the Streamlit app
+if __name__ == '__main__':
+    app()