-
Notifications
You must be signed in to change notification settings - Fork 3.5k
LightningCLI instantiator receives values applied by instantiation links to set in hparams #20777
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: master
Are you sure you want to change the base?
LightningCLI instantiator receives values applied by instantiation links to set in hparams #20777
Conversation
save_hyperparameters
consistent for CLI and hardcoded training for custom python objects
#20432
@@ -5,7 +5,7 @@ | |||
matplotlib>3.1, <3.9.0 | |||
omegaconf >=2.2.3, <2.4.0 | |||
hydra-core >=1.2.0, <1.4.0 | |||
jsonargparse[signatures] >=4.27.7, <=4.35.0 | |||
jsonargparse[signatures] >=4.39.0, <4.40.0 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think the tests that fail is because the newer version of jsonargparse is not installed.
Please check the failing GPU stand-alone test:
|
Looking at the logs, in |
What does this PR do?
Fixes
save_hyperparameters
when used throughLightningCLI
(feature introduced in #18105) so that target values of links applied for instantiation are included in hparams.Fixes #20311
Fixes #20147
Also adds an official way to disable #18105 (parameter
load_from_checkpoint_support
), which might be desired by some people, e.g #20432, #20726Before submitting
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
📚 Documentation preview 📚: https://pytorch-lightning--20777.org.readthedocs.build/en/20777/