Share the training step output data via ClosureResult

[carmocca]

What does this PR do?

Remove the need to use deepcopy.

Fixes #8821

Follow-up PRs

• Treat everything returned during manual_optimization as an extra.
  • Finite checks: will be removed after detect_anomaly and in any case your training_step would have failed already.
  • Running loss: No longer updated after extract manual optimization loop #9266 (comment) and potentially to be removed with [RFC] Remove {running,accumulated}_loss #9372
• Warn when extras are returned but training_epoch_end is not implemented.
• Split out the result dataclasses, as the behavior would be different in automatic compared to manual:

@dataclass
class OutputResult:
    all attributes optional

    defines utility methods to avoid duplication

class ManualResult(OutputResult):
    def from_taining_step_output(): ...
class ClosureResult(OutputResult):
    def from_taining_step_output(): ...

Notes:

# manual optimization `training_step` possible return formats:
(1): None # most common (doesn't skip)
(2): a_loss # Tensor. what does this do?
(3): {'loss': ...}  # same as (2)
(4): {'loss': ..., 'anything': ...}  # includes an extra
(5): {'loss': None, 'anything': ...}  #  you dont want to return a loss but need the extra
(6): {'loss': ..., 'hiddens': ...}
(7): {'loss': None, 'hiddens': ...}

Does your PR introduce any breaking changes? If yes, please list them.

None

Before submitting

• Was this discussed/approved via a GitHub issue? (not for typos and docs)
• Did you read the contributor guideline, Pull Request section?
• Did you make sure your PR does only one thing, instead of bundling different changes together?
• Did you make sure to update the documentation with your changes? (if necessary)
• Did you write any new necessary tests? (not for typos and docs)
• Did you verify new and existing tests pass locally with your changes?
• Did you list all the breaking changes introduced by this pull request?
• Did you update the CHANGELOG? (not for typos, docs, test updates, or internal minor changes/refactorings)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing make sure you have read Review guidelines. In short, see the following bullet-list:

• Is this pull request ready for review? (if not, please submit in draft mode)
• Check that all items from Before submitting are resolved
• Make sure the title is self-explanatory and the description concisely explains the PR
• Add labels and milestones (and optionally projects) to the PR so it can be classified

[tchaton]

Looks great !

[codecov]

Codecov Report

Merging #9349 (65ab8d9) into master (c963bf6) will decrease coverage by 0%.
The diff coverage is 97%.

@@          Coverage Diff           @@
##           master   #9349   +/-   ##
======================================
- Coverage      93%     93%   -0%     
======================================
  Files         179     179           
  Lines       14927   14915   -12     
======================================
- Hits        13868   13854   -14     
- Misses       1059    1061    +2     

[leezu]

@carmocca this PR fixes a bug that was also backported/introduced in the 1.4.6 release (#9239). Is it possible to backport this PR as well to 1.4?

[carmocca]

We didn't cherry pick this because it builds on top of several refactors that landed before and aren't in the bug-fix branch.

A custom fix for it in the bug-fix branch could be explored but it's not a high priority right now.