Meta-Learning for Natural Language Understanding under Continual Learning Framework
This work addresses continual learning for natural language understanding, which is incremental as it applies existing meta-learning methods to a new framework.
The paper tackled the problem of natural language understanding under continual learning by implementing model-agnostic meta-learning (MAML) and Online aware Meta-learning (OML) objectives, validating them on SuperGLUE and GLUE benchmarks.
Neural network has been recognized with its accomplishments on tackling various natural language understanding (NLU) tasks. Methods have been developed to train a robust model to handle multiple tasks to gain a general representation of text. In this paper, we implement the model-agnostic meta-learning (MAML) and Online aware Meta-learning (OML) meta-objective under the continual framework for NLU tasks. We validate our methods on selected SuperGLUE and GLUE benchmark.