Adding support for relational operations #783
Closed
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
facebook-github-bot
added
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 20, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) Vectorization is not done yet. - this maybe fixed in future iterations (4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Differential Revision: D46770449 fbshipit-source-id: 32d5e4dc1165791f41a9d55124db488fbb826cdd
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
58fba27 to
b344066
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 21, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) Vectorization is not done yet. - this maybe fixed in future iterations (4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Differential Revision: D46770449 fbshipit-source-id: 5f4a727787532708c390a2728aedf1d0dcaf7f37
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
b344066 to
8640858
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 21, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) Vectorization is not done yet. - this maybe fixed in future iterations (4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Differential Revision: D46770449 fbshipit-source-id: 90afab845afd5d33fa817a5ef79c0e7baf41b999
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
8640858 to
b5326b8
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 21, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) Vectorization is not done yet. - this maybe fixed in future iterations (4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Differential Revision: D46770449 fbshipit-source-id: 26cde876d117f956f5a46401b7a3a17ba50331a7
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
b5326b8 to
dd65419
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 21, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) Vectorization is not done yet. - this maybe fixed in future iterations (4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Differential Revision: D46770449 fbshipit-source-id: 6916799ec57356d1f2ca5e886551a79172421a94
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
dd65419 to
990c60a
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 26, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Reviewed By: aakhundov Differential Revision: D46770449 fbshipit-source-id: 238a304e6bd261363dac92cb31552af9e568d3eb
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
990c60a to
6a2ecce
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 26, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Reviewed By: aakhundov Differential Revision: D46770449 fbshipit-source-id: e264ade7f0a39a73311caa29951209ed1b38edff
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
6a2ecce to
9d164d0
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
AlbertDachiChen
added a commit
to AlbertDachiChen/AITemplate
that referenced
this pull request
Jun 27, 2023
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Reviewed By: aakhundov Differential Revision: D46770449 fbshipit-source-id: e8dd9453f4940936b0da7ce7fc383aa0eafc3288
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
9d164d0 to
dfb19b8
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
Summary: Pull Request resolved: facebookincubator#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Reviewed By: aakhundov Differential Revision: D46770449 fbshipit-source-id: c985355b556af536543fc701053bcc57c71b8084
AlbertDachiChen
force-pushed
the
export-D46770449
branch
from
dfb19b8 to
10b02e2
Compare
|
This pull request was exported from Phabricator. Differential Revision: D46770449 |
|
This pull request has been merged in 3c3b256. |
Ivorchu
pushed a commit
to Ivorchu/AITemplate-NM-Pruning-Archive
that referenced
this pull request
Jul 30, 2025
Summary: Pull Request resolved: facebookincubator/AITemplate#783 As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html There are multiple constraints: (1) Type promotions are not supported. (2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations (3) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations Reviewed By: aakhundov Differential Revision: D46770449 fbshipit-source-id: ed58c8ff0148b389dc80f76fe6a4413f21c36a2f
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.
Summary:
As titled, this diffs adds supports for relational operations (e.g. ge, le, gt, lt, eq, ne). The expected behavior is to match equivalent operators in Torch, e.g. https://pytorch.org/docs/stable/generated/torch.ge.html
There are multiple constraints:
(1) Type promotions are not supported.
(2) Broadcast is not supported yet, so it expects tensor of the same shape - this maybe fixed in future iterations
(3) Vectorization is not done yet. - this maybe fixed in future iterations
(4) It uses the relational operators, i.e, >=, <, > rather than comparison function that supports different level of precision, e.g. https://docs.nvidia.com/cuda/cuda-math-api/group__CUDA__MATH____HALF__COMPARISON.html - this maybe fixed in future iterations
Differential Revision: D46770449