Call configure_model from LightningCLI

[ioangatop]

What does this PR do?

Fixes #19110

Before submitting

• Was this discussed/agreed 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 minor internal changes/refactors)

PR review

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

Reviewer checklist

• 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

---

📚 Documentation preview 📚: https://pytorch-lightning--19111.org.readthedocs.build/en/19111/

[ioangatop]

Looking at the following, probably the solution needs to be adjusted:

pytorch-lightning/src/lightning/pytorch/trainer/call.py

Lines 99 to 109 in 520c1e4

	def _call_configure_model(trainer: "pl.Trainer") -> None:
	# legacy hook
	if is_overridden("configure_sharded_model", trainer.lightning_module):
	with trainer.strategy.model_sharded_context():
	_call_lightning_module_hook(trainer, "configure_sharded_model")
	# we don't normally check for this before calling the hook. it is done here to avoid instantiating the context
	# managers
	if is_overridden("configure_model", trainer.lightning_module):
	with trainer.strategy.tensor_init_context(), trainer.strategy.model_sharded_context(), trainer.precision_plugin.module_init_context(): # noqa: E501
	_call_lightning_module_hook(trainer, "configure_model")

The issue is that the above excepts the trainer which is not yet initialised.

Now the configure_model will be called twice (like if we have the model initialisation both in __init__ and in the configure_model) 🤔 - maybe it works but idk if this behaviour is desirable

However, it does get called here directly:

pytorch-lightning/src/lightning/pytorch/core/saving.py

Lines 161 to 162 in 3cbc14e

	if is_overridden("configure_model", obj):
	obj.configure_model()

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79%. Comparing base (eb082b2) to head (14ed93f).
Report is 19 commits behind head on master.

❗ There is a different number of reports uploaded between BASE (eb082b2) and HEAD (14ed93f). Click for more details.

HEAD has 753 uploads less than BASE

Flag	BASE (eb082b2)	HEAD (14ed93f)
cpu	192	24
python	24	3
lightning_fabric	38	0
pytest	99	0
python3.10	48	6
lightning	121	15
python3.11	48	6
gpu	3	0
python3.12.7	72	9
pytorch2.1	24	6
pytorch_lightning	36	9
pytest-full	96	24
pytorch2.2.2	12	3
pytorch2.3	12	3
pytorch2.5	12	3
pytorch2.4.1	12	3
pytorch2.5.1	12	3
pytorch2.6	12	3

Additional details and impacted files

@@            Coverage Diff            @@
##           master   #19111     +/-   ##
=========================================
- Coverage      88%      79%     -9%     
=========================================
  Files         267      264      -3     
  Lines       23364    23310     -54     
=========================================
- Hits        20467    18355   -2112     
- Misses       2897     4955   +2058     

[ioangatop]

@Borda thanks for the ✅ - is there anyone we can tag that can review any maybe merge/resolve the issue? 🙏

[gitguardian]

️✅ There are no secrets present in this pull request anymore.

If these secrets were true positive and are still valid, we highly recommend you to revoke them.
While these secrets were previously flagged, we no longer have a reference to the
specific commits where they were detected. Once a secret has been leaked into a git
repository, you should consider it compromised, even if it was deleted immediately.
Find here more information about risks.

---

^{_{🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.}}

[ioangatop]

@Borda any update for this? thank you

[Borda]

@lantiga mind have a look? :)

[stale]

This pull request has been automatically marked as stale because it has not had recent activity. It will be closed in 7 days if no further activity occurs. If you need further help see our docs: https://lightning.ai/docs/pytorch/latest/generated/CONTRIBUTING.html#pull-request or ask the assistance of a core contributor here or on Discord. Thank you for your contributions.