Update README.md

Fix readme

Author: s-gupta

README.md

@@ -48,6 +48,7 @@ Code for detecting visual concepts in images.
   # were not complete. v2 tar ball fixes this.
   wget ftp://ftp.cs.berkeley.edu/pub/projects/vision/im2cap-cvpr15b/trained-coco.v2.tgz && tar -xf trained-coco.v2.tgz 
   ```
+
 #### Demo ####
 Look at ``demo.ipynb`` for a demo in IPython notebook.