search

yeyupiaoling / VoiceprintRecognition-Pytorch

This project uses a variety of advanced voiceprint recognition models such as EcapaTdnn, ResNetSE, ERes2Net, CAM++, etc. It is not excluded that more models will be supported in the future. At the same time, this project also supports MelSpectrogram, Spectrogram data preprocessing methods
Apache License 2.0
777 stars 124 forks source link

关于测试 #32

Closed EdenJin20171503024 closed 2 years ago

EdenJin20171503024 commented 2 years ago

我发现您的这个项目中训练时，在验证的时候用的时test list，但是在eval的时候也是用的test去测试模型的效果，是不是在train的时候验证的时候用dev集比较好

yeyupiaoling commented 2 years ago

只是验证测试时没必要这么分。

jzz990415 commented 2 years ago

能问一下这个模型的话是有哪篇论文吗，想研究一下

yeyupiaoling commented 2 years ago

你搜索下ecapa_tdnn