SYCL. Fix multiclass for sycl iGPUs #11292
Draft
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
|
I will look into refactoring the code after the 3.0 release. Dispatching f64 seems to be quite intrusive for the codebase. We use f64 a lot. |
|
Does iGPU with sycl have a software-simulated f64 implementation? It seems quite restrictive to not have a standard floating point. You are not going to get very far without f64 for traditional machine learning algorithms. (R uses f64 as default, Python float is f64, sklearn, numpy also uses f64 as default). |
|
This is less of iGPUs but particular generation of ARC systems, with newer Batlemage generation having fp64 support. |
|
That's great news, we can drop the workarounds for old generation. |
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Outdated suggestions cannot be applied.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.
The PR #11029 generalizes calculations of multiclass objectives for sycl. But some sycl devices (integrated graphics) don't support fp64 calculations. This fix adds dispatching to gradient calculations. And returns the basic xgboost functional to iGPUs.