[Feature] tversky loss

[MengzhangLI]

Adding tversky loss.

Not too much difference with original dice_loss.py, just adding weight on false positives and false negatives.

[codecov]

Codecov Report

Merging #634 (188e7e1) into master (3e70d93) will increase coverage by 0.34%.
The diff coverage is 98.75%.

❗ Current head 188e7e1 differs from pull request most recent head 5ba5acd. Consider uploading reports for the commit 5ba5acd to get more accurate results

@@            Coverage Diff             @@
##           master     #634      +/-   ##
==========================================
+ Coverage   85.57%   85.91%   +0.34%     
==========================================
  Files         101      104       +3     
  Lines        5226     5360     +134     
  Branches      842      862      +20     
==========================================
+ Hits         4472     4605     +133     
  Misses        583      583              
- Partials      171      172       +1     

Flag	Coverage Δ
unittests	85.91% <98.75%> (+0.34%)	⬆️

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

Impacted Files	Coverage Δ
mmseg/datasets/pipelines/transforms.py	97.12% <ø> (ø)
mmseg/models/utils/__init__.py	100.00% <ø> (ø)
mmseg/models/backbones/vit.py	85.87% <66.66%> (-0.80%)	⬇️
mmseg/models/decode_heads/__init__.py	100.00% <100.00%> (ø)
mmseg/models/decode_heads/setr_mla_head.py	100.00% <100.00%> (ø)
mmseg/models/decode_heads/setr_up_head.py	100.00% <100.00%> (ø)
mmseg/models/losses/__init__.py	100.00% <100.00%> (ø)
mmseg/models/losses/tversky_loss.py	100.00% <100.00%> (ø)
mmseg/models/necks/__init__.py	100.00% <100.00%> (ø)
mmseg/models/necks/mla_neck.py	100.00% <100.00%> (ø)
... and 4 more

---

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 3e70d93...5ba5acd. Read the comment docs.

[MengzhangLI]

Because creating new PR, so this old version is closed.