Skip to content

Remove @cached decorator for TriangularLinearOperator.to_dense#51

Merged
gpleiss merged 2 commits intocornellius-gp:mainfrom
Balandat:no_cached_tlop_todense
Feb 27, 2023
Merged

Remove @cached decorator for TriangularLinearOperator.to_dense#51
gpleiss merged 2 commits intocornellius-gp:mainfrom
Balandat:no_cached_tlop_todense

Conversation

@Balandat
Copy link
Collaborator

This seems unnecessary and also causes issues with pickling: cornellius-gp/gpytorch#2264 (comment)

This seems unnecessary and also causes issues with pickling: cornellius-gp/gpytorch#2264 (comment)
@esantorella
Copy link
Collaborator

My gut instinct is that this is right, since I've seen the same dense tensor getting cached to multiple places, including via this function. If this causes any performance degradation, hopefully the BoTorch tutorials runs and the tracking I'm adding to them would pick that up (meta-pytorch/botorch#1695).

@gpleiss gpleiss enabled auto-merge (squash) February 27, 2023 22:18
@gpleiss gpleiss merged commit 92f7e33 into cornellius-gp:main Feb 27, 2023
@Balandat Balandat deleted the no_cached_tlop_todense branch April 13, 2023 19:35
Balandat added a commit to Balandat/linear_operator that referenced this pull request May 3, 2023
gpleiss pushed a commit that referenced this pull request May 3, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants