Robust Tickets Can Transfer Better: Drawing More Transferable Subnetworks in Transfer Learning
This work addresses the deployment challenge of large models for edge devices, representing an incremental improvement to the lottery ticket hypothesis.
The paper tackles the problem of deploying large pretrained models on resource-constrained devices by proposing a transfer learning pipeline that uses robust subnetworks, achieving enhanced accuracy-sparsity trade-offs across diverse downstream tasks.
Transfer learning leverages feature representations of deep neural networks (DNNs) pretrained on source tasks with rich data to empower effective finetuning on downstream tasks. However, the pretrained models are often prohibitively large for delivering generalizable representations, which limits their deployment on edge devices with constrained resources. To close this gap, we propose a new transfer learning pipeline, which leverages our finding that robust tickets can transfer better, i.e., subnetworks drawn with properly induced adversarial robustness can win better transferability over vanilla lottery ticket subnetworks. Extensive experiments and ablation studies validate that our proposed transfer learning pipeline can achieve enhanced accuracy-sparsity trade-offs across both diverse downstream tasks and sparsity patterns, further enriching the lottery ticket hypothesis.