added micro average option for torch metrics

[razmikmelikbekyan]

What does this PR do?

Added Micro average option for IoU.

Fixes #826

Before submitting

• Was this discussed/approved via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure to update the docs?
• Did you write any new necessary tests?

PR review

Anyone in the community is free to review the PR once the tests have passed.
If we didn't discuss your PR in Github issues there's a high chance it will not be merged.

Did you have fun?

Make sure you had fun coding 🙃

[codecov]

Codecov Report

Merging #874 (1008e88) into master (567f0ea) will increase coverage by 15%.
The diff coverage is 96%.

❗ Current head 1008e88 differs from pull request most recent head 5435cec. Consider uploading reports for the commit 5435cec to get more accurate results

@@           Coverage Diff           @@
##           master   #874     +/-   ##
=======================================
+ Coverage      71%    86%    +15%     
=======================================
  Files         181    181             
  Lines        7993   8004     +11     
=======================================
+ Hits         5681   6895   +1214     
+ Misses       2312   1109   -1203     

[razmikmelikbekyan]

@Borda Please have a look. After initial review, I will remove all the commented-out lines, but there might be some incompatibility related issues since I was going to remove the reduction parameter fully.

[justusschock]

implementation-wise it looks good to me. Just not sure how to proceed with different arguments for different metrics. IMO it should all be called reduction, but not sure how to handle different arguments. Will we have separate helper functions or just one function that does it all and some others just for input validation?

[Borda]

@razmikmelikbekyan how is it going here? 🐰

[razmikmelikbekyan]

@razmikmelikbekyan how is it going here? 🐰

@Borda Honestly said I am a bit confused about the next steps. I am not sure that I will have enough time right now to refactor all the other metrics and so on and I do not understand what exactly needs to be done in this particular PR.

Should we merge it or we want to change name back to reduction and so one (in my opinion name average_type is more relevant and adapted by the community for this kind of metrics than reduction)

[Borda]

Honestly said I am a bit confused about the next steps.

@stancld mind help with this PR? :)

[justusschock]

Hey @razmikmelikbekyan ,

Very sorry for the long return time. The decisions made offline with @Borda and @SkafteNicki were that we want to get this to merge now and that we will deal with the naming inconsistencies while doing the major refactor (#1001 ).
That being said, do you think, there is a chance that you could look into the failing tests?

[razmikmelikbekyan]

Hey @razmikmelikbekyan ,

Very sorry for the long return time. The decisions made offline with @Borda and @SkafteNicki were that we want to get this to merge now and that we will deal with the naming inconsistencies while doing the major refactor (#1001 ). That being said, do you think, there is a chance that you could look into the failing tests?

hey @justusschock, yep, I will have a look soon.