Search results

Why softmax instead of sigmoid is used?

Status: Open.

#15 In clcarwin/focal_loss_pytorch;

· Podidiving opened on Apr 9, 2023

Negative examples

Status: Open.

#14 In clcarwin/focal_loss_pytorch;

· ShJacub opened on Dec 26, 2022

Loss occurs nan!

Status: Open.

#13 In clcarwin/focal_loss_pytorch;

· YuYue26 opened on Jan 5, 2022

alpha value is not consistent with paper?

Status: Open.

#10 In clcarwin/focal_loss_pytorch;

· eedavidwu opened on May 30, 2020

Normalisation & mean over classes

Status: Open.

#8 In clcarwin/focal_loss_pytorch;

· karanchahal opened on Feb 18, 2019

Shouldn't it be log_sigmoid instead of softmax ?

Status: Open.

#7 In clcarwin/focal_loss_pytorch;

· karanchahal opened on Feb 18, 2019

the problem of reshape

Status: Open.

#6 In clcarwin/focal_loss_pytorch;

· weixia1 opened on Dec 22, 2018

shall the pt bp?

Status: Open.

#5 In clcarwin/focal_loss_pytorch;

· sillytheif opened on Dec 21, 2018

if I set alpha the program crash

Status: Open.

#4 In clcarwin/focal_loss_pytorch;

· stephane-archer opened on Oct 31, 2018

What are the best parameters?

Status: Open.

#3 In clcarwin/focal_loss_pytorch;

· stephane-archer opened on Sep 6, 2018

FocalLoss vs CrossEntropyLoss

Status: Open.

#2 In clcarwin/focal_loss_pytorch;

· nationalflag opened on Dec 14, 2017

NameError: name 'long' is not defined

Status: Open.

#1 In clcarwin/focal_loss_pytorch;

· pengchengma opened on Aug 28, 2017