-
Notifications
You must be signed in to change notification settings - Fork 142
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.
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
Fx fixes #402
Fx fixes #402
Conversation
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Thanks a lot for the PR, it looks good! |
Not sure, there is a test failure, but I can't tell if it's a problem or if it actually fixes earlier issue : |
If you can investigate a bit this bug it's cool, otherwise we can also just comment it for now |
I guess it's just because now some assertions are inside an if statement |
Instead of wrapping things in conditionals, I think you could replace See https://pytorch.org/docs/stable/generated/torch._assert.html. This is cleaner and doesn't change the behavior of the functions under tracing. |
Signed-off-by: Boris Fomitchev <bfomitchev@nvidia.com>
Right; I have used Lisp's suggestion now :) |
@mariogeiger : I guess it's ready for merge now :) |
Good job |
Changes to make TensorProduct classes FX-traceable
Description
Those changes are partial and would only allow subset of classes used in DiffDock, pass FX symbolic_trace().
jit_script_fx needs to be set to False for this to work, too.
Motivation and Context
Resolves: #???
How Has This Been Tested?
Checklist: