Closed db091756 closed 1 month ago
Was this resolved when #746 was merged?
Was this resolved when #746 was merged?
We reverted the line of code that caused the issue so it is as fast as it was previously. But it is still worth understanding where the issue is. I believe it is something to do with computing the hessian but not looked into it fully
What's the new feature?
Changing how enforce 2-dimensionality in the
score_matching.py
code slows downSteinThinning
very significantly. Track down this bug. (See https://github.com/gchq/coreax/pull/746#issuecomment-2321449604)What value does this add?
Consistent dimensionality handling without slowdown!
Is there an alternative you've considered?
No response
Additional context
No response