Gumbel-Matrix Routing for Flexible Multi-task Learning

arxiv(2019)

引用 0|浏览33
暂无评分
摘要
This paper proposes a novel per-task routing method for multi-task applications. Multi-task neural networks can learn to transfer knowledge across different tasks by using parameter sharing. However, sharing parameters between unrelated tasks can hurt performance. To address this issue, we advocate the use of routing networks to learn flexible parameter sharing, where each group of parameters is shared with a different subset of tasks in order to better leverage tasks relatedness. At the same time, it is known that routing networks are notoriously hard to train. We propose the Gumbel-Matrix routing: a novel multi-task routing method, designed to learn fine-grained patterns of parameter sharing. The routing is learned jointly with the model parameters by standard back-propagation thanks to the Gumbel-Softmax trick. When applied to the Omniglot benchmark, the proposed method reduces the state-of-the-art error rate by 17%.
更多
查看译文
关键词
learning,gumbel-matrix,multi-task
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要