[tune] make tune.with_parameters() work with the class API
#14532
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
krfricke
commented
Mar 8, 2021
Comment on lines
+300
to
+312
| if inspect.isclass(trainable): | ||
| # Class trainable | ||
| keys = list(kwargs.keys()) | ||
|
|
||
| class _Inner(trainable): | ||
| def setup(self, config): | ||
| setup_kwargs = {} | ||
| for k in keys: | ||
| setup_kwargs[k] = parameter_registry.get(prefix + k) | ||
| super(_Inner, self).setup(config, **setup_kwargs) | ||
|
|
||
| _Inner.__name__ = trainable_name | ||
| return _Inner |
There was a problem hiding this comment.
This is the main change here
krfricke
commented
Mar 8, 2021
Comment on lines
+313
to
+345
| else: | ||
| # Function trainable | ||
| use_checkpoint = detect_checkpoint_function(trainable, partial=True) | ||
| keys = list(kwargs.keys()) | ||
|
|
||
| def inner(config, checkpoint_dir=None): | ||
| fn_kwargs = {} | ||
| if use_checkpoint: | ||
| default = checkpoint_dir | ||
| sig = inspect.signature(trainable) | ||
| if "checkpoint_dir" in sig.parameters: | ||
| default = sig.parameters["checkpoint_dir"].default \ | ||
| or default | ||
| fn_kwargs["checkpoint_dir"] = default | ||
|
|
||
| for k in keys: | ||
| fn_kwargs[k] = parameter_registry.get(prefix + k) | ||
| trainable(config, **fn_kwargs) | ||
|
|
||
| inner.__name__ = trainable_name | ||
|
|
||
| # Use correct function signature if no `checkpoint_dir` parameter | ||
| # is set | ||
| if not use_checkpoint: | ||
|
|
||
| def _inner(config): | ||
| inner(config, checkpoint_dir=None) | ||
|
|
||
| _inner.__name__ = trainable_name | ||
|
|
||
| if hasattr(trainable, "__mixins__"): | ||
| _inner.__mixins__ = trainable.__mixins__ | ||
| return _inner |
There was a problem hiding this comment.
This is just copy/pasted from the old file
richardliaw
approved these changes
Mar 9, 2021
richardliaw
reviewed
Mar 9, 2021
Co-authored-by: Richard Liaw <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are these changes needed?
tune.with_parameters()currently only works with function trainables, but it is actually a nice way to pass data to class trainables, too.This questions came up several times in the past, e.g. https://discuss.ray.io/t/improper-run-not-string-nor-trainable/1135/8
Related issue number
Checks
scripts/format.shto lint the changes in this PR.