Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Use PyTorch Functionalization Pass #30

Closed
narendasan opened this issue Apr 1, 2020 · 3 comments · Fixed by #183
Closed

Use PyTorch Functionalization Pass #30

narendasan opened this issue Apr 1, 2020 · 3 comments · Fixed by #183
Labels
component: lowering Issues re: The lowering / preprocessing passes feature request New feature or request priority: high
Milestone
v0.1.0

Comments

@narendasan
Copy link
Collaborator

Having converters for inplace operators (add_) is potentially unsafe. We should use the functionalization pass in the lowering phase to switch these ops to their functional versions.
@narendasan narendasan added feature request New feature or request component: lowering Issues re: The lowering / preprocessing passes priority: high labels Apr 1, 2020
@narendasan narendasan added this to the v0.0.2 milestone Apr 1, 2020
@narendasan
Copy link
Collaborator Author

https://github.com/pytorch/pytorch/blob/cd5d7a34b8802cf120451182fd17a662efec644c/torch/csrc/jit/passes/create_functional_graphs.h

@github-actions
Copy link

github-actions bot commented Jul 4, 2020

This issue has not seen activity for 30 days, Remove stale label or comment or this will be closed in 5 days

@github-actions github-actions bot closed this as completed Jul 9, 2020
@narendasan narendasan reopened this Jul 9, 2020
@github-actions
Copy link

github-actions bot commented Aug 9, 2020

This issue has not seen activity for 30 days, Remove stale label or comment or this will be closed in 5 days

narendasan added a commit that referenced this issue Sep 4, 2020
@narendasan
feat(//core/lowering): Add functionalization pass to replace implace
90a9ed6 
operators

Closes: #30

Signed-off-by: Naren Dasan <[email protected]>
Signed-off-by: Naren Dasan <[email protected]>
@narendasan narendasan mentioned this issue Sep 4, 2020
Merged
6 tasks
andi4191 pushed a commit to andi4191/Torch-TensorRT that referenced this issue Oct 30, 2020
@narendasan
feat(//core/lowering): Add functionalization pass to replace implace
2e275f9 
operators

Closes: pytorch#30

Signed-off-by: Naren Dasan <[email protected]>
Signed-off-by: Naren Dasan <[email protected]>
frank-wei pushed a commit that referenced this issue Jun 4, 2022
@alexbeloi
[fx][acc_ops] add acc_ops.gather and acc_ops.index_select and shape i…
730645d 
…nference (#30)

Summary:
Pull Request resolved: https://github.com/pytorch/fx2trt/pull/30

as title

Reviewed By: 842974287

Differential Revision: D34874487

fbshipit-source-id: 86039d0f1269d879983977c65fe8fbe0a8bc1421
@lanluo-nvidia lanluo-nvidia mentioned this issue Aug 7, 2024
Merged
7 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
component: lowering Issues re: The lowering / preprocessing passes feature request New feature or request priority: high
Projects
None yet
Milestone
v0.1.0
Development

Successfully merging a pull request may close this issue.

feat(//core/lowering): Add functionalization pass to replace inplace operators pytorch/TensorRT
1 participant
@narendasan