This work represents a significant step forward in adapting sequential stochastic model-based convex optimization methods to the much more realistic mini-batch setting. Congratulations on the nice work! As a minor comment, I agree with the reviewer that "mini-batch" might be better than "parallel" in the title.