Early return in _recursive_build if obj is a Tensor #125639
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/125639
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit bcc4944 with merge base f155ed6 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@pytorchbot rebase |
|
@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here |
|
Successfully rebased |
guilhermeleobas
changed the title
|
I'm pretty sure you have to clone this at minimum... look at the code above. |
| if isinstance(obj, Tensor) and obj.ndim <= 1: | ||
| return obj.detach().to(dtype=scalarType, device="cpu", copy=True).view(()) | ||
| elif isinstance(obj, Tensor): | ||
| return obj.detach().to(dtype=scalarType, device="cpu", copy=True) |
There was a problem hiding this comment.
I'm not sure you'll necessarily end up with the right shape without an appropriate view. Specifically:
>>> torch.tensor([torch.tensor(0)])
tensor([0])
>>> torch.tensor([torch.tensor([0])])
tensor([0])
>>> torch.tensor([torch.tensor([[0]])])
tensor([0])
>>> torch.tensor([torch.tensor([[[0]]])])
tensor([0])
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Fix issue pytorch#125551 Pull Request resolved: pytorch#125639 Approved by: https://github.com/ezyang
Stack from ghstack (oldest at bottom):
Fix issue #125551
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @chenyang78 @kadeng @chauhang