Update open_clip.md (#1485)

Add import requests line

docs/hub/open_clip.md

@@ -34,6 +34,7 @@ Once loaded, you can encode the image and text to do [zero-shot image classifica
 ```py
 import torch
 from PIL import Image
+import requests
 
 url = 'http://images.cocodataset.org/val2017/000000039769.jpg'
 image = Image.open(requests.get(url, stream=True).raw)