search

Tramac / Fast-SCNN-pytorch

A PyTorch Implementation of Fast-SCNN: Fast Semantic Segmentation Network
Apache License 2.0
381 stars 93 forks source link

the loss decline too slow #51

Open DOVEXQ opened 1 year ago

DOVEXQ commented 1 year ago

Hi, I use this project to train Cityscapes, use adam optim，lr is 3e-3,but the loss decline is too slow ,how can i solve this problem' Thanks

Alohomora0130 commented 1 year ago

Have you solve this problem? I meet the same situation.