[tabular] Improve NN_TORCH runtime estimate #4247
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
|
suzhoum
approved these changes
Jun 5, 2024
|
Job PR-4247-4341b4b is done. |
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Outdated suggestions cannot be applied.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.
Issue #, if available:
Description of changes:
When running NN_TORCH in a
rayprocess, the first batch includes significant time overhead to initialize torch. This leads to the mainline time requirement estimate to be vastly pessimistic as the data becomes larger because it multiplies this time overhead by the number of batches in an epoch. This caused extreme estimates for large datasets, skipping the neural network training entirely in many cases where it would have been useful.The fix avoids using the first batch as part of the runtime estimate of future batches. This allows the
v2time estimate to be far more accurate.Example on the adult income dataset:
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.