Fetching the paper…

On Good Practices for Task-Specific Distillation of Large Pretrained Visual Models · Around