Model Uncertainty-Aware Knowledge Amalgamation for Pre-Trained Language Models
This addresses the need to reduce computational costs and environmental impacts in NLP by efficiently merging specialized models, though it is an incremental advancement in model reuse techniques.
The paper tackles the problem of reusing multiple fine-tuned pre-trained language models (PLMs) without human annotations by proposing a knowledge amalgamation (KA) paradigm, resulting in a versatile student model that achieves substantial improvements over baselines on benchmark datasets.
As many fine-tuned pre-trained language models~(PLMs) with promising performance are generously released, investigating better ways to reuse these models is vital as it can greatly reduce the retraining computational cost and the potential environmental side-effects. In this paper, we explore a novel model reuse paradigm, Knowledge Amalgamation~(KA) for PLMs. Without human annotations available, KA aims to merge the knowledge from different teacher-PLMs, each of which specializes in a different classification problem, into a versatile student model. The achieve this, we design a Model Uncertainty--aware Knowledge Amalgamation~(MUKA) framework, which identifies the potential adequate teacher using Monte-Carlo Dropout for approximating the golden supervision to guide the student. Experimental results demonstrate that MUKA achieves substantial improvements over baselines on benchmark datasets. Further analysis shows that MUKA can generalize well under several complicate settings with multiple teacher models, heterogeneous teachers, and even cross-dataset teachers.