Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Errors in loading the pretrained model resnet50 and Low cityscape performance #56

Open
MMCheng opened this issue Aug 4, 2020 · 0 comments

Comments

@MMCheng
Copy link

MMCheng commented Aug 4, 2020

1. Errors in loading the pretrained model resnet50.
What is the kernel size of the layers: module.layer0.0.weight, module.layer0.3.weight, module.layer0.6.weight

2. Low cityscape performance with pretrained pspnet50. mIoU/mAcc/allAcc 0.6444/0.7284/0.9410
I set the kernel size to 3 in the layers: layer1.0.conv1.weight and layer1.0.downsample.0.
I trained with 4 RTX 2080t GPUs and the set the batch_size to 12.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant