Hyper-Learning for Gradient-Based Batch Size Adaptation

Re-assign community