torch/ao #2029
torch/ao #2029
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2029
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit ac50759 with merge base ef10f34 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
This pull request was exported from Phabricator. Differential Revision: D72435827 |
| if TORCH_VERSION_AT_LEAST_2_5: | ||
| model = torch.export.export_for_training(model, example_inputs).module() | ||
| model = torch.export.export_for_training( | ||
| model, example_inputs, strict=True |
There was a problem hiding this comment.
is strict False by default now? do we need to add this everywhere?
There was a problem hiding this comment.
strict=False is not by default yet, but it will be. I don't think you need to add this everywhere as in vast majority of use cases, the end graphs should be the same.
There was a problem hiding this comment.
I see, then why do we add strict=True here if the end model will be the same?
There was a problem hiding this comment.
I think @gmagogsfm was just doing a code mod change and this is just for preserving the old behaviour?
Summary: Pull Request resolved: pytorch#2029 Reviewed By: ydwu4 Differential Revision: D72435827
b5d8a53 to
ac50759
Compare
|
This pull request was exported from Phabricator. Differential Revision: D72435827 |
Reviewed By: ydwu4
Differential Revision: D72435827