[MetaSchedule] Evo Independence from TaskScheduler #11590
Merged
Conversation
This file contains hidden or 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
Per discussion with @Kathryn-cat, we realized that the current API design could be verbose if we only want to tune a single task, in which case a dummy task scheduler still needs to be established to supply `EvolutionarySearch` with proper `CostModel` and `Database`. This PR fixes this UX issue.
Contributor
|
I tested on local and it looks good! Perhaps add a hint showing that |
Member
Author
|
Thank you @Kathryn-cat for the valuable feedback! Will do! |
Member
Author
|
@Kathryn-cat Hey I improved the error message with clear instructions of what to do. Let me know if it looks good :-) |
Kathryn-cat
approved these changes
Jun 7, 2022
Contributor
Kathryn-cat
left a comment
Kathryn-cat
left a comment
There was a problem hiding this comment.
Files in this PR look good! Perhaps initialize_with_tune_context can be removed from some test files in the future and be marked for internal use only.
Hzfengsy
approved these changes
Jun 7, 2022
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
Per discussion with @Kathryn-cat, we realized that the current API
design could be verbose if we only want to tune a single task, in which
case a dummy task scheduler still needs to be established to supply
EvolutionarySearchwith properCostModelandDatabase. This PRfixes this UX issue.