Kunyuan Ding

http://www.dkyblog.com

Student of Beijing University of Posts and Telecommunications

Results 2 issues of


                                            Kunyuan Ding

Can the model refuse to answer when it cannot guarantee the result？

Thanks for your work，可否在输出结果时加入判断，若置信度较低或不曾在训练集见过类似的问题就让模型拒绝回答问题？

what's the difference between self(x) and self.model(x) in training_step

features = self(x) I wonder what's the different between self(x) and use the network directly such as self.model(x)?