Support soft cross entropy loss

[Junjun2016]

Support soft cross entropy loss

[codecov[bot]]

Codecov Report

Merging #252 (a168850) into master (2bd51e6) will increase coverage by 0.16%. The diff coverage is 96.66%.

@@            Coverage Diff             @@
##           master     #252      +/-   ##
==========================================
+ Coverage   84.43%   84.59%   +0.16%     
==========================================
  Files          90       91       +1     
  Lines        4361     4421      +60     
  Branches      688      697       +9     
==========================================
+ Hits         3682     3740      +58     
- Misses        537      538       +1     
- Partials      142      143       +1     

Flag	Coverage Δ
unittests	84.59% <96.66%> (+0.16%)	:arrow_up:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmseg/models/losses/soft_cross_entropy_loss.py	96.61% <96.61%> (ø)
mmseg/models/losses/__init__.py	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 2bd51e6...db1c408. Read the comment docs.

[yiminglin-ai]

How is the performance compared with standard (hard) CrossEntropy?