Asynchronous Convergence in Multi-Task Learning via Knowledge Distillation from Converged Tasks.
Weiyi LuSunny RajagopalanPriyanka NigamJaspreet SinghXiaodi SunYi XuBelinda ZengTrishul ChilimbiPublished in: NAACL-HLT (Industry Papers) (2022)
Keyphrases
- multi task learning
- multi task
- multiple tasks
- multitask learning
- learning tasks
- gaussian processes
- previously learned
- transfer learning
- learning problems
- high order
- theoretical analysis
- expert systems
- knowledge base
- learning algorithm
- knowledge representation
- data sets
- learning models
- prior knowledge
- reinforcement learning