arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2110.02432cs.CLcs.AIcs.LG
收藏

KNOT:基于最优传输的知识蒸馏方法求解NLP任务

KNOT: Knowledge Distillation using Optimal Transport for Solving NLP Tasks

Singapore University of Technology and Design(新加坡科技设计大学);Nanyang Technological University(南洋理工大学)

Rishabh Bhardwaj, Tushar Vaidya, Soujanya Poria

摘要:

我们提出了一种新的方法,知识蒸馏使用最优传输(KNOT),提取自然语言语义知识从多个教师网络到学生网络。KNOT旨在训练(全局)学生模型,其方法是在学生模型不能访问教师模型的参数或训练数据的约束下,学习将其在标签上的分配概率分布的最优传输成本最小化为由(局部)教师模型预测的概率的加权和。为了评估知识转移的质量,我们引入了一个新的度量标准,语义距离(SD),它测量预测和地面真实标签分布之间的语义接近度。所提出的方法显示,在三个NLP任务中,全局模型的SD性能比基线有所改善,同时在标准精度和F1指标上与基于熵的蒸馏相媲美。与这项工作有关的实现可在https://github.com/declare-lab/KNOT上公开获得。

英文摘要:

We propose a new approach, Knowledge Distillation using Optimal Transport (KNOT), to distill the natural language semantic knowledge from multiple teacher networks to a student network. KNOT aims to train a (global) student model by learning to minimize the optimal transport cost of its assigned probability distribution over the labels to the weighted sum of probabilities predicted by the (local) teacher models, under the constraints, that the student model does not have access to teacher models' parameters or training data. To evaluate the quality of knowledge transfer, we introduce a new metric, Semantic Distance (SD), that measures semantic closeness between the predicted and ground truth label distributions. The proposed method shows improvements in the global model's SD performance over the baseline across three NLP tasks while performing on par with Entropy-based distillation on standard accuracy and F1 metrics. The implementation pertaining to this work is publicly available at: https://github.com/declare-lab/KNOT.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑