Update handler.py

handler.py

@@ -1,3 +1,4 @@
+import requests
 from typing import Dict, List, Any
 from PIL import Image
 import torch
@@ -16,16 +17,17 @@ class EndpointHandler():
         self.model = self.model.to(device)
 
     def __call__(self, data: Dict[str, Any]) -> Dict[str, Any]:
-        images = data.get("images")
+        image_urls = data.get("images")
+        
         # Check if images is None or empty and handle it appropriately
-        if not images:
-            return {"captions": []} 
+        if not image_urls:
+            return {"captions": [], "error": "No images provided"} 
         
         # Default to "a photography of" if texts not provided
-        texts = data.get("texts", ["a photography of"] * len(images)) 
+        texts = data.get("texts", ["a photography of"] * len(image_urls)) 
         
         try:
-            raw_images = [Image.open(BytesIO(_img)).convert("RGB") for _img in images]
+            raw_images = [Image.open(requests.get(url, stream=True).raw).convert("RGB") for url in image_urls]
             processed_inputs = [
                 self.processor(img, txt, return_tensors="pt") for img, txt in zip(raw_images, texts)
             ]
@@ -43,4 +45,4 @@ class EndpointHandler():
         except Exception as e:
             # Handle or log the exception and optionally return an error message
             print(f"Error during processing: {str(e)}")
-            return {"error": str(e)}
+            return {"captions": [], "error": str(e)}