Fetching the paper…

APT: Adaptive Pruning and Tuning Pretrained Language Models for Efficient Training and Inference · Around