Add MPS regression test for ExactGP to track Apple Silicon backend support#2748
Draft
dascoder wants to merge 3 commits into
Draft
Add MPS regression test for ExactGP to track Apple Silicon backend support#2748dascoder wants to merge 3 commits into
dascoder wants to merge 3 commits into
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
With the recent fixes coming upstream from PyTorch (specifically the native linalg.cholesky support in PR #172536), Apple Silicon (MPS) is technically capable of running native ExactGP models without triggering math crashes. However, GPyTorch currently lacks a dedicated hardware test to track MPS compatibility.
This PR adds a minimal test_exact_gp_mps regression test to the hardware triad (CPU/CUDA/MPS) to prevent future silent regressions.
Note on Performance Bottlenecks
While this test passes (proving the math is stable on M1/M2 chips), it surfaces a PyTorch-level fallback warning indicating that aten::_cholesky_solve_helper is missing from the Metal backend, causing a silent CPU fallback. I will add proposal to bypass this overhead in Issue #2209