SKDBERT: Compressing BERT via Stochastic Knowledge Distillation

Abstract In this paper, we propose Stochastic Knowledge Distillation (SKD) to obtain compact BERT-style language model dubbed SKDBERT. In each distillation iteration, SKD samples a teacher model from a pre-defined teacher team, which consists of multiple teacher models with multi-level capacities, to transfer knowledge into student model in an one-to-one manner. Sampling distribution plays an important role in SKD. We heuristically present three types of sampling distributions to assign appropriate probabilities for multi-level teacher models. SKD has two advantages: 1) it can preserve the diversities of multi-level teacher models via stochastically sampling single teacher model in each distillation iteration, and 2) it can also improve the efficacy of knowledge distillation via multi-level teacher models when large capacity gap exists between the teacher model and the student model. Experimental results on GLUE benchmark show that SKDBERT reduces the size of a BERT model by 40% while retaining 99.5% performances of language understanding and being 100% faster.