Wrap constant arg in XLAExportInterpreter #6460
Merged
Conversation
This file contains 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
JackCaoG
reviewed
Feb 2, 2024
JackCaoG
approved these changes
Feb 3, 2024
amithrm
pushed a commit
to amithrm/xla
that referenced
this pull request
Mar 1, 2024
bhavya01
pushed a commit
that referenced
this pull request
Apr 22, 2024
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Outdated suggestions cannot be applied.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.
If the op spec expects a
Tensorinput, we wrap the python primitive type to a torch.tensor. The dtype for tensor respectstorch.get_default_dtype.Without this wrapping, the python float will be wrapped before it enters dispatcher, and it doesn't respect the global default dtype.
With this change,
aten.sub(1.0, x)andaten.sub(torch.tensor(1.0), x)will be the same, when they enters the LTC.It's still unknown where the scalar constant is wrapped into a tensor in upstream. It should be before it enters dispatcher (By investigation and syncing with PyTorch team).