Update app.py

app.py

@@ -1,58 +1,26 @@
-import streamlit as st
+# Dog v Cat script
+__all__ = ['is_cat','learn','classify_image','categories','image','label','examples','intf']
+
+# cell
 from fastai.vision.all import *
+import gradio as gr
 
-def is_cat(x): 
-    return x[0].isupper()
+def is_cat(x): return x[0].isupper()
 
-# Load the pre-trained model
-model_path = "model.pkl"  # Replace with the path to your model file
-learn = load_learner(model_path)
+# cell
+learn = load_learner('model.pkl')
 
-# Define categories
-categories = ['Dog', 'Cat']
+# cell
+categories = ('Dog','Cat')
 
-# Function to classify an image
 def classify_image(img):
-    pred, idx, probs = learn.predict(img)
-    return dict(zip(categories, map(float, probs)))
-
-# Streamlit app
-st.title("Dog vs. Cat Classifier")
-
-# Upload an image
-uploaded_image = st.file_uploader("Upload an image", type=["jpg", "jpeg", "jfif", "png"])
-
-if uploaded_image is not None:
-    # Display the uploaded image
-    st.image(uploaded_image, caption="Uploaded Image", use_column_width=True)
-
-    # Make predictions
-    image = PILImage.create(uploaded_image)
-    predictions = classify_image(image)
-
-    # Display the prediction
-    st.subheader("Prediction:")
-    for category, probability in predictions.items():
-        st.write(f"{category}: {probability:.2f}")
-
-# Example images
-st.sidebar.title("Example Images")
-example_images = {
-    "Dog": "dog.jfif",
-    "Cat": "cat.jfif"
-}
-
-selected_example = st.sidebar.selectbox("Select an Example Image", list(example_images.keys()))
-
-if selected_example:
-    selected_image_path = example_images[selected_example]
-    st.image(selected_image_path, caption=selected_example, use_column_width=True)
+    pred,idx,probs = learn.predict(img)
+    return dict(zip(categories, map(float,probs)))
 
-    # Make predictions for the example image
-    example_image = PILImage.create(selected_image_path)
-    example_predictions = classify_image(example_image)
+# cell
+image = gr.inputs.Image(shape=(192,192))
+label = gr.outputs.Label()
+examples = ['dog.jfif','cat.jfif']
 
-    # Display the prediction
-    st.sidebar.subheader("Prediction:")
-    for category, probability in example_predictions.items():
-        st.sidebar.write(f"{category}: {probability:.2f}")
+intf = gr.Interface(fn=classify_image, inputs=image, outputs= label, examples=examples)
+intf.launch(inline=False)