Support S32/U32 indices for BWD embedding and index ops & Neuron implicit downcast#8450
Closed
rpsilva-aws wants to merge 17 commits intopytorch:masterfrom
Closed
Support S32/U32 indices for BWD embedding and index ops & Neuron implicit downcast#8450rpsilva-aws wants to merge 17 commits intopytorch:masterfrom
rpsilva-aws wants to merge 17 commits intopytorch:masterfrom
Conversation
Collaborator
|
is this PR ready for review @rpsilva-aws? |
Collaborator
Author
|
@miladm Yes, but there seem to be some issues with the test infra - I need to validate with the TPU docker. I can ping once done/fixed. |
Collaborator
|
Can just ping me again when the "Build and test" has passed |
Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
…ter/tpu-recipies repo for reference model implementations. (pytorch#8412)
Collaborator
Author
|
Unfortunately, part of this PR has a dependency on pytorch/pytorch#142160. Instead, creating 2 PRs:
|
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
10 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
In this PR, we extend tensor operations to allow S32 indices. This follows suits with other operations, in order to add flexibility and potentially performance benefits for accelerator backends. Reference for embedding dense bwd: https://github.com/pytorch/pytorch/blob/main/aten/src/ATen/native/Embedding.cpp#L117
I re-structure a good portion of the tests to validate both types.
In addition, we also re-introduce the implicit downcasting for Neuron S64/U64 types, since the Neuron compiler does not support 64 bits.