[data] [streaming] Improve progress bar rendering on large scale jobs #33952
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
ericl
commented
Mar 30, 2023
| @@ -405,17 +405,6 @@ def map_batches( | |||
| To learn more about writing functions for :meth:`~Dataset.map_batches`, read | |||
| :ref:`writing user-defined functions <transform_datasets_writing_udfs>`. | |||
|
|
|||
| .. tip:: | |||
There was a problem hiding this comment.
Remove overly spammy / incorrect tips
There was a problem hiding this comment.
Why it's incorrect though? We no longer want to move usage to higher-level APIs like preprocessor ?
There was a problem hiding this comment.
It's kind of the opposite of ray-project/enhancements#25.
I don't think this was ever correct in the first place. It's not really good for Data users to be told to use preprocessor when they may not need it.
jianoaix
approved these changes
Mar 30, 2023
| @@ -405,17 +405,6 @@ def map_batches( | |||
| To learn more about writing functions for :meth:`~Dataset.map_batches`, read | |||
| :ref:`writing user-defined functions <transform_datasets_writing_udfs>`. | |||
|
|
|||
| .. tip:: | |||
There was a problem hiding this comment.
Why it's incorrect though? We no longer want to move usage to higher-level APIs like preprocessor ?
|
@zhe-thoughts we could consider merging this for 2.4, don't have a strong opinion. |
ericl
added
tests-ok
c21
approved these changes
Mar 31, 2023
|
OK then let's wait until we unfreeze master. |
8 tasks
bveeramani
added a commit
that referenced
this pull request
Aug 29, 2024
…47393) In each Ray Data scheduling step, Ray Data launches tasks until it can't launch any more. To ensure that the progress bar updates frequently, #33952 made it such that Ray Data can only launch 50 tasks per scheduling step. However, this cause performance issues for large workloads. This PR fixes the issue by removing the limit on number of tasks launched while still updating the progress bar frequently.
ujjawal-khare
pushed a commit
to ujjawal-khare-27/ray
that referenced
this pull request
Oct 12, 2024
…ay-project#47393) In each Ray Data scheduling step, Ray Data launches tasks until it can't launch any more. To ensure that the progress bar updates frequently, ray-project#33952 made it such that Ray Data can only launch 50 tasks per scheduling step. However, this cause performance issues for large workloads. This PR fixes the issue by removing the limit on number of tasks launched while still updating the progress bar frequently. Signed-off-by: ujjawal-khare <[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?
For large jobs, the progress bar can be jerky since we do not limit the number of events processed per progress bar update.