Model compression using knowledge distillation with integrated gradients

Open in new window