Skip to content

The probability caculation of negative class in 'Label smooth.py' #19

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.

Sign up for GitHub

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

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

The probability caculation of negative class in 'Label smooth.py' #19

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
8ad2ae5
Update label_smooth.py
kaijieshi7 Feb 24, 2021
24bd671
Merge pull request #1 from kaijieshi7/kaijieshi7-patch-1
kaijieshi7 Feb 24, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions label_smooth.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,7 +36,7 @@ def forward(self, logits, label):
ignore = label.eq(self.lb_ignore)
n_valid = ignore.eq(0).sum()
label[ignore] = 0
lb_pos, lb_neg = 1. - self.lb_smooth, self.lb_smooth / num_classes
lb_pos, lb_neg = 1. - self.lb_smooth, self.lb_smooth / (num_classes - 1)
lb_one_hot = torch.empty_like(logits).fill_(
lb_neg).scatter_(1, label.unsqueeze(1), lb_pos).detach()

Expand All @@ -61,7 +61,7 @@ class LSRCrossEntropyFunctionV2(torch.autograd.Function):
def forward(ctx, logits, label, lb_smooth, lb_ignore):
# prepare label
num_classes = logits.size(1)
lb_pos, lb_neg = 1. - lb_smooth, lb_smooth / num_classes
lb_pos, lb_neg = 1. - lb_smooth, lb_smooth / (num_classes - 1)
label = label.clone().detach()
ignore = label.eq(lb_ignore)
n_valid = ignore.eq(0).sum()
Expand Down