How to adjust the learning rate in 4 RTX 2080ti？

CSAILVision / semantic-segmentation-pytorch

Pytorch implementation for Semantic Segmentation/Scene Parsing on MIT ADE20K dataset

http://sceneparsing.csail.mit.edu/

BSD 3-Clause "New" or "Revised" License

4.96k stars 1.1k forks source link

How to adjust the learning rate in 4 RTX 2080ti？ #207

Closed BangguWu closed 4 years ago

BangguWu commented 5 years ago

I have 8 RTX 2080TI, but when I am training ade20k-resnet50dilated-ppm_deepsup.yaml, it will appear out of memory. But I can train it on 4 RTX 2080TI, How should I adjust the hyperparameters like learning rate or others?

hangzhaomit commented 4 years ago

Try smaller batch size, performance might get worse though.