Lightning-AI / Lightning-AI/lightning-thunder

`frexp` backward returns garbage values for `float64` inputs of larger shapes

Open
#1,991 0 comments 0 reactions 1 assignee View on GitHub

@IvanYashchuk is already working on this.

Since Apr 30, 2025.

autograd operators testing
Dominant language
Python
Stars
1.5k
Forks
121
PR merge metrics
No merged PRs in 30d

Description

`test_vjp_correctness` fails with the following error:

```python
E AssertionError: Scalars are not close!
E
E Expected 303979.2360823408 but got 54382.22813814061.
E Absolute difference: 249597.00794420022 (up to 1e-07 allowed)
E Relative difference: 0.8210988722814945 (up to 1e-07 allowed)
```
This error is due to the following test case:

https://github.com/Lightning-AI/lightning-thunder/blob/a8c063676cbecad7c196dcfe4fbeacf85775c9c5/thunder/tests/opinfos.py#L485-L489

Discovered in https://github.com/Lightning-AI/lightning-thunder/pull/1896#discussion_r2014481184

cc @mruberry

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.