Paper “Repeated Random Sampling for Minimizing the Time-to-Accuracy of Learning” has been accepted to ICLR 2024.