Adding non-bert-based models to Open Intent Detection

[mxnno]

Hi, in your paper you wrote: "for example, the primary backbone is the pre-trained BERT (Devlin et al., 2019) model, and it supports adding new bert-based models with different downstream tasks." But I would like to add other models for the Open Intent Detection part aswell, which are not bert-based, like Prototypical Networks, in order to compare more Open Intent Detection methods within this framework. So is this possible?

[HanleiZhang]

Hi, you can add different models in the directory 'backbones'. Specifically, you need first config the basic model in the 'base.py' (such as pattern == 'bert') and then add new models with your basic backbones (such as 'bert.py').