Update app.py

app.py

@@ -3,6 +3,7 @@ import os
 import io
 from PIL import Image
 import requests
+import base64
 import warnings
 import gradio as gr
 
@@ -24,23 +25,25 @@ if not endpoint_url:
 def generate_caption(image):
     """
     Sends an image to the Hugging Face Inference Endpoint for caption generation.
+    Uses base64 encoding for compatibility.
     :param image: An image in PIL format.
     :return: Generated caption or error message.
     """
     try:
         headers = {"Authorization": f"Bearer {hf_api_key}"}
         
-        # Convert the PIL image to a binary stream in JPEG format
+        # Convert the image to RGB and encode it in base64
         buffered = io.BytesIO()
         image = image.convert("RGB")  # Ensure the image is in RGB mode
         image.save(buffered, format="JPEG")
         buffered.seek(0)
+        image_base64 = base64.b64encode(buffered.read()).decode("utf-8")
         
-        # Create the appropriate payload for the API
-        files = {"file": ("image.jpg", buffered, "image/jpeg")}
+        # Prepare the JSON payload
+        payload = {"inputs": image_base64}
         
         # Make the POST request to the endpoint
-        response = requests.post(endpoint_url, headers=headers, files=files)
+        response = requests.post(endpoint_url, headers=headers, json=payload)
         
         if response.status_code == 200:
             return response.json().get("generated_text", "No caption generated.")
@@ -88,3 +91,4 @@ if __name__ == "__main__":
     # Launch the Gradio demo
     demo.launch()
 
+