The reviewers agree that the paper is solid and make novel contributions to understanding federated learning and meta-learning.