Back to Search
Start Over
AdaKD: Dynamic Knowledge Distillation of ASR models using Adaptive Loss Weighting
- Publication Year :
- 2024
-
Abstract
- Knowledge distillation, a widely used model compression technique, works on the basis of transferring knowledge from a cumbersome teacher model to a lightweight student model. The technique involves jointly optimizing the task specific and knowledge distillation losses with a weight assigned to them. Despite these weights playing a crucial role in the performance of the distillation process, current methods provide equal weight to both losses, leading to suboptimal performance. In this paper, we propose Adaptive Knowledge Distillation, a novel technique inspired by curriculum learning to adaptively weigh the losses at instance level. This technique goes by the notion that sample difficulty increases with teacher loss. Our method follows a plug-and-play paradigm that can be applied on top of any task-specific and distillation objectives. Experiments show that our method performs better than conventional knowledge distillation method and existing instance-level loss functions.
- Subjects :
- Computer Science - Machine Learning
Computer Science - Artificial Intelligence
Subjects
Details
- Database :
- arXiv
- Publication Type :
- Report
- Accession number :
- edsarx.2405.08019
- Document Type :
- Working Paper