Few Shot Dialog State Tracking using Meta-learning
Source code for our "D-REPTILE" paper at EACL 2021: Saket Dingliwal, Bill Gao, Sanchit Agarwal, Tagyoung Chung, and Dilek Hakkani-tur. "Few Shot Dialogue State Tracking using Meta-learning" EACL (2021). If you use the code, please cite the paper:
@article{}
Work in Progress
Please stay tuned for the complete codebase and the specifications of the model and the hyper-parameters.
Installation
Pipeline
The code consists of the implementation of D-REPTILE and the baseline (Naive Fine-Tuning) using the base model STARC. The implementation train and evaluate the categorical and non-categorical slots of each domain seperately. The directory structure is explained below
cat
non-cat