Add cummin operation for tensors across backends #3821
Closed
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
Introduces the cummin (cumulative minimum) operation for both float and int tensors, with implementations for all supported backends (ndarray, tch, candle, cubecl, fusion, router). Updates the tensor API, backend traits, and IR to support cummin, and adds comprehensive tests and documentation. The autodiff backend panics for cummin as proper gradient support is not yet implemented.
2 tasks
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## main #3821 +/- ##
==========================================
- Coverage 64.27% 64.21% -0.06%
==========================================
Files 1114 1115 +1
Lines 133659 133919 +260
==========================================
+ Hits 85904 85999 +95
- Misses 47755 47920 +165 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
2 tasks
laggui
requested changes
Oct 3, 2025
Comment on lines
+1669
to
+1675
| // Cummin backward pass requires scatter_add which is not yet implemented | ||
| // The gradient should only flow to the first occurrence of each minimum value | ||
| panic!( | ||
| "Cummin is not supported for autodiff backend. \ | ||
| Proper implementation requires scatter_add operation." | ||
| ); | ||
| } |
There was a problem hiding this comment.
Same comment as cummax: tensor.scatter applies the sum reduction (scatter_add equivalent).
We need to improve this expected behavior discrepancy at the tensor API level. select_assign also performs a sum 😅
|
Merged into #3819 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Introduces the cummin (cumulative minimum) operation for both float and int tensors, with implementations for all supported backends (ndarray, tch, candle, cubecl, fusion, router). Updates the tensor API, backend traits, and IR to support cummin, and adds comprehensive tests and documentation. The autodiff backend panics for cummin as proper gradient support is not yet implemented.
Pull Request Template
Checklist
cargo run-checkscommand has been executed.Related Issues/PRs
Changes
Testing
Added 5 comprehensive tests in burn-tensor/src/tests/ops/cummin.rs