[dtensor][random] use torch.uint64 as the seed/offset tensor dtype to avoid overflow #141532
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
… avoid overflow [ghstack-poisoned]
This was referenced Nov 26, 2024
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/141532
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 5 Unrelated FailuresAs of commit 74c82f7 with merge base 6a22cae ( NEW FAILURE - The following job has failed:
FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
oncall: distributed
…or dtype to avoid overflow" cc H-Huang awgu kwen2501 wanchaol fegin fduwjj wz337 wconstab d4l3k c-p-i-o [ghstack-poisoned]
…or dtype to avoid overflow" **Summary** DTensor RNG code raises error if the seed passed in is beyong `torch.int64` range (e.g. `torch.tensor([2**64-1])` raises error). The solution is to specify the `dtype=torch.uint64` in the `torch.tensor()` call. cc H-Huang awgu kwen2501 wanchaol fegin fduwjj wz337 wconstab d4l3k c-p-i-o [ghstack-poisoned]
wconstab
approved these changes
Nov 26, 2024
Contributor
wconstab
left a comment
wconstab
left a comment
There was a problem hiding this comment.
glad you found this, i noticed the error once but not consistently, and didn't track down where it came from.
Skylion007
reviewed
Nov 27, 2024
torch/distributed/tensor/_random.py
Outdated
|
|
||
| seed_tensor = (self.rng_states[name])[0:8] | ||
| offset_tensor = torch.tensor([offset]).view(torch.uint8) | ||
| offset_tensor = torch.tensor([offset], dtype=torch.uint64).view(torch.uint8) |
Collaborator
There was a problem hiding this comment.
Isn't this immediately cast to uint8?
Contributor
There was a problem hiding this comment.
offset comes in as a uint64 essentially. we're using a tensor to convert it to its constituent bytes because this is the format used by torch's RNG state. the initial dtype=uint64 is needed to prevent overflow warnings mainly.
Collaborator
|
Oh I see, this is just to hide an error. |
…or dtype to avoid overflow" **Summary** DTensor RNG code raises error if the seed passed in is beyong `torch.int64` range (e.g. `torch.tensor([2**64-1])` raises error). The solution is to specify the `dtype=torch.uint64` in the `torch.tensor()` call. cc H-Huang awgu kwen2501 wanchaol fegin fduwjj wz337 wconstab d4l3k c-p-i-o [ghstack-poisoned]
XilunWu
added
ciflow/trunk
Contributor
Author
|
@pytorchbot merge -f " |
Collaborator
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
GeorgeWigley
pushed a commit
to graphcore/pytorch-fork
that referenced
this pull request
Nov 29, 2024
… avoid overflow (pytorch#141532) **Summary** DTensor RNG code raises error if the seed passed in is beyong `torch.int64` range (e.g. `torch.tensor([2**64-1])` raises error). The solution is to specify the `dtype=torch.uint64` in the `torch.tensor()` call. Pull Request resolved: pytorch#141532 Approved by: https://github.com/wconstab ghstack dependencies: pytorch#141731, pytorch#141220, pytorch#141223
pobin6
pushed a commit
to pobin6/pytorch
that referenced
this pull request
Dec 5, 2024
… avoid overflow (pytorch#141532) **Summary** DTensor RNG code raises error if the seed passed in is beyong `torch.int64` range (e.g. `torch.tensor([2**64-1])` raises error). The solution is to specify the `dtype=torch.uint64` in the `torch.tensor()` call. Pull Request resolved: pytorch#141532 Approved by: https://github.com/wconstab ghstack dependencies: pytorch#141731, pytorch#141220, pytorch#141223
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Stack from ghstack (oldest at bottom):
Summary
DTensor RNG code raises error if the seed passed in is beyong
torch.int64range (e.g.torch.tensor([2**64-1])raises error). The solution is to specify thedtype=torch.uint64in thetorch.tensor()call.cc @H-Huang @awgu @kwen2501 @wanchaol @fegin @fduwjj @wz337 @wconstab @d4l3k @c-p-i-o