-
Notifications
You must be signed in to change notification settings - Fork 525
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
[torch-mlir] provide FX traced graph importer for sparse tensors #2817
Conversation
Note that we are waiting for actual FX traced graph support for sparse tensors. For details see pytorch/pytorch#117188 Until then, however, we provide this clever importer that builds the FX traced graph for for the dense case and then puts a sparse annotation back on the parameters. With import test.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I wasn't quite ready to accept contributions to this part yet. Let me synchronize the downstream fork tomorrow and make sure we don't get into a merge conflict.
I suspect conflicts will be very light, since I am only touching a few tensor type building parts. But let me know when you are ready. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should be fine but I'm just trying to juggle less. I'm juggling a downstream pr and almost have the spice flowing to not need the fork. Give me a few hours
Of course. Take your time! Just reply here when you are comfortable moving forward. |
I've gone ahead and taken steps to unfork the code from the downstream (iree-org/iree#16276). Will take a release cycle to percolate but in the meantime, we're fine with this minor difference. |
Thanks for your flexibility moving this forward quickly! |
Note that we are waiting for actual FX traced graph support for sparse tensors. For details see
pytorch/pytorch#117188
Until then, however, we provide this clever importer that builds the FX traced graph for for the dense case and then puts a sparse annotation back on the parameters.
With import test.