Update app.py

app.py

@@ -1,34 +1,26 @@
 import streamlit as st
+import requests
 
-# Load the model
-model = load_model('model.h5')
+st.title("Image Captioning App")
 
-# Define the input and output data
-input_data = st.text_input('Enter input data:')
-output_data = st.text_input('Enter output data:')
+# Create a file uploader widget
+uploaded_file = st.file_uploader("Choose an image file", type="image/*")
 
-# Define the prediction function
-def predict(input_data):
-    # Preprocess the input data
-    input_data = preprocess_input(input_data)
-    
-    # Make the prediction
-    output_data = model.predict(input_data)
-    
-    # Postprocess the output data
-    output_data = postprocess_output(output_data)
-    
-    return output_data
+# If a file is uploaded, send a POST request to the API endpoint
+if uploaded_file is not None:
+    # Read the image file as a base64 encoded string
+    image_data = uploaded_file.read()
+    image_data = base64.b64encode(image_data)
 
-# Create the Streamlit app
-st.title('Image Captioning App')
-st.write('Enter input data:')
-st.write(input_data)
-st.write('Enter output data:')
-st.write(output_data)
+    # Send a POST request to the API endpoint with the image data as the payload
+    response = requests.post("https://ashrafb-salesforce-blip-image-captioning-base.hf.space/run/predict", json={
+        "data": [
+            image_data,
+        ]
+    })
 
-# Add a button to make the prediction
-if st.button('Make Prediction'):
-    output_data = predict(input_data)
-    st.write('Predicted output:')
-    st.write(output_data)
+    # Extract the generated caption from the response
+    caption = response.json()["data"][0]
+
+    # Display the generated caption
+    st.write(caption)