Fix _generate_seed_sequence sampling#21399
Merged
justusschock merged 3 commits intoLightning-AI:masterfrom Dec 10, 2025
Merged
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## master #21399 +/- ##
=======================================
- Coverage 87% 87% -0%
=======================================
Files 269 269
Lines 23804 23809 +5
=======================================
+ Hits 20626 20628 +2
- Misses 3178 3181 +3 |
justusschock
approved these changes
Dec 10, 2025
bhimrazy
pushed a commit
to bhimrazy/pytorch-lightning
that referenced
this pull request
Jan 30, 2026
lantiga
pushed a commit
that referenced
this pull request
Jan 30, 2026
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Fixes #20412
Currently the
_generate_seed_sequencefunction does not produce unique seeds. The reason for this is that small base seeds (e.g. 0-999) after left-shifting by 32 bits have all zeros in the lower 32 bits. The LCG algorithm currently being used with c=1, then produces seeds that all havelower_32_bits=1which in turn causestorch.manual_seed()to treat them identically since Pytorch primarily uses the lower bits.Implements a simple hash mixing that ensures even small seeds produce well-distributed bit patterns across the full 64-bit space, preventing the current seed collision we are seeing.
Before submitting
PR review
Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:
Reviewer checklist
📚 Documentation preview 📚: https://pytorch-lightning--21399.org.readthedocs.build/en/21399/