[fx2trt] support for ne, logical_not, logical_and #75444
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
Summary: as titled 1. support logical_and, logical_not 2. replace eq,gt,lt with python operator in acc_ops due to the fact that torch op needs input to be torch.Tensor but python op does not 3. add more test cases 4. add individual ne op without using combination of existing ops since there are limitations. For ex, in lowering it to equal+logical_not. It will fail the last test case in test_ne.py. The failure reason is that logical_not needs the input to be a tensor. Also we can not use equal+operator.not since not is not tracable in FX with the error "symbolically traced variables cannot be used as inputs to control flow" We also can not use equal+operator.invert since operator.invert(True)=-2 Test Plan: buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_ne buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_logical_and buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_unary_ops Reviewed By: 842974287 Differential Revision: D35232917 fbshipit-source-id: 60368a6f73071403394a369ed94e319a9dd86c29
Contributor
🔗 Helpful links
💊 CI failures summary and remediationsAs of commit 51e9ec3 (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Please report bugs/suggestions to the (internal) Dr. CI Users group. |
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D35232917 |
facebook-github-bot
pushed a commit
that referenced
this pull request
Apr 7, 2022
Summary: Pull Request resolved: #75444 as titled 1. support logical_and, logical_not 2. replace eq,gt,lt with python operator in acc_ops due to the fact that torch op needs input to be torch.Tensor but python op does not 3. add more test cases 4. add individual ne op without using combination of existing ops since there are limitations. For ex, in lowering it to equal+logical_not. It will fail the last test case in test_ne.py. The failure reason is that logical_not needs the input to be a tensor. Also we can not use equal+operator.not since not is not tracable in FX with the error "symbolically traced variables cannot be used as inputs to control flow" We also can not use equal+operator.invert since operator.invert(True)=-2 (Note: this ignores all push blocking failures!) Test Plan: buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_ne buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_logical_and buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_unary_ops Reviewed By: 842974287 Differential Revision: D35232917 fbshipit-source-id: d4601a6883c977caa263f67b9db86cbc862d4780
Contributor
|
Hey @frank-wei. |
frank-wei
pushed a commit
to pytorch/TensorRT
that referenced
this pull request
Jun 4, 2022
Summary: X-link: pytorch/pytorch#75444 as titled 1. support logical_and, logical_not 2. replace eq,gt,lt with python operator in acc_ops due to the fact that torch op needs input to be torch.Tensor but python op does not 3. add more test cases 4. add individual ne op without using combination of existing ops since there are limitations. For ex, in lowering it to equal+logical_not. It will fail the last test case in test_ne.py. The failure reason is that logical_not needs the input to be a tensor. Also we can not use equal+operator.not since not is not tracable in FX with the error "symbolically traced variables cannot be used as inputs to control flow" We also can not use equal+operator.invert since operator.invert(True)=-2 (Note: this ignores all push blocking failures!) Reviewed By: 842974287 Differential Revision: D35232917 fbshipit-source-id: d4601a6883c977caa263f67b9db86cbc862d4780
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
Summary:
as titled
Also we can not use equal+operator.not since not is not tracable in FX with the error "symbolically traced variables cannot be used as inputs to control flow"
We also can not use equal+operator.invert since operator.invert(True)=-2
Test Plan:
buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_ne
buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_logical_and
buck test mode/dev-nosan deeplearning/trt/fx2trt_oss/test/converters:test_unary_ops
Reviewed By: 842974287
Differential Revision: D35232917