[tune] Store sync config/checkpoint config in experiment, trial #29019
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
added 7 commits
October 3, 2022 15:04
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
krfricke
requested review from
sven1977,
gjoliver,
avnishn,
ArturNiederfahrenhorst,
smorad,
maxpumperla and
kouroshHakha
as code owners
|
Overall looks good! Thanks for the cleanup. Just need to see the tests are passing.. |
added 3 commits
October 4, 2022 14:32
Signed-off-by: Kai Fricke <[email protected]>
Signed-off-by: Kai Fricke <[email protected]>
…/checkpoint-retry-env
xwjiang2010
approved these changes
Oct 5, 2022
gjoliver
approved these changes
Oct 5, 2022
WeichenXu123
pushed a commit
to WeichenXu123/ray
that referenced
this pull request
Dec 19, 2022
…project#29019) This is some clean-up required for future changes to the syncing/checkpointing behavior. At the moment we pass single attributes of these configs to the Experiment class, and then subsequently to the Trial class, from which it is passed on to the trainable. If we extend the configurability in the future (e.g. provide fallback mechanisms in the checkpoint config, or make retry wait times configurable in the sync config), we would have to add more and more attributes to these intermediate classes. Instead, we should just pass and store the full config. As a next follow-up, we can pass these configs to the Trainable. Signed-off-by: Kai Fricke <[email protected]> Signed-off-by: Weichen Xu <[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.
Signed-off-by: Kai Fricke [email protected]
Why are these changes needed?
This is some clean-up required for future changes to the syncing/checkpointing behavior. At the moment we pass single attributes of these configs to the Experiment class, and then subsequently to the Trial class, from which it is passed on to the trainable. If we extend the configurability in the future (e.g. provide fallback mechanisms in the checkpoint config, or make retry wait times configurable in the sync config), we would have to add more and more attributes to these intermediate classes. Instead, we should just pass and store the full config.
As a next follow-up, we can pass these configs to the Trainable.
Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.