Are the models you have supported in the code the model you get the results in the paper?

fidler-lab / polyrnn-pp-pytorch

PyTorch training/tool code for Polygon-RNN++ (CVPR 2018)

704 stars 106 forks source link

Are the models you have supported in the code the model you get the results in the paper? #41

Closed LetsGoFir closed 4 years ago

LetsGoFir commented 4 years ago

I find that the epoch is very small, like 'ggnn_epoch8_step21000.pth', but the epochs you have set in the config is about 200.

amlankar commented 4 years ago

Hey, yes the models released with the code are in fact better than the ones reported in the paper. We have set epochs to 200, but usually stop training much earlier (early stopping)

LetsGoFir commented 4 years ago

Thanks