-
-
Notifications
You must be signed in to change notification settings - Fork 8.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
CPU optimizations - 'hist' method #5104
Comments
I would love to see the performance optimization back and sorry for the oversight. Feel free to ping me if you need any help. |
For now, a script that we can run is okay, so that performance results can be re-produced. I plan to design and build an automatic system for performance benchmarking. |
All PRs are merged, final results:
Current master is is slightly slower on MNIST data set than it was before, but it can be covered in the next PRs Thank you all who helped me to refactor the optimizations to have them in the master again (@hcho3, @trivialfis, @RAMitchell ). |
Moved discussion from PR #5008 to the issue. After the PR - performance drop on CPU is observed. I prepared a plan how to establish optimizations from PR #4529 in master branch again.
I tried to split large changes to separate PRs:
Expected timelines for creation the all PRs - EOY.
The text was updated successfully, but these errors were encountered: