that our implementation will be a widely used tool for embedding convex optimization problems in end-to-end learning