Fast Catch-Up, Late Switching: Optimal Batch Size Scheduling via Functional Scaling Laws

arXiv:2602.14208v2 Announce Type: replace-cross Abstract: Batch size scheduling (BSS) plays a critical role in large-scale deep learning training, influencing both optimization dynamics and computational efficiency. Yet, its theoretical foundations remain poorly understood. In...

🔗 Read more: https://arxiv.org/abs/2602.14208

#News #Policy #AI #Academic