Question about utils/evaluation.py

maum-ai / voicefilter

Unofficial PyTorch implementation of Google AI's VoiceFilter system

http://swpark.me/voicefilter

1.09k stars 227 forks source link

Question about utils/evaluation.py #13

Open lawlict opened 5 years ago

lawlict commented 5 years ago

Hello @seungwonpark , thank you greatly for your work! I notice that the utils/evalluation.py has a "break" in the loop of test dataloader. That is, in the evaluation process, only the first case generated from test dataloader is taken account into computing test loss and test SDR. Could this raise problems like #5 and #9 ?

Look forward to your response.

seungwonpark commented 5 years ago

Hi, @lawlict

Yes, our current SDR calculation process may cause SDR value to look strange since it evaluates only once.

bigcash commented 3 years ago

@lawlict from dataloader.py you can get the batchsize of testloader have only 1, so evalluation.py has a "break" may be ok