Support torch.compile
skingorz opened this issue ยท 2 comments
skingorz commented
๐ Feature
I hope the repo support torch.compile.
Motivation
As a key feature of Pytorch 2.0, I hope USB can also support torch.compile.
Additional context
I have try to add the following code (line 390-392) to train.py. But when the second epoch ends and evals the model ,
eval_dict = algorithm.evaluate('eval')
it always occurs an assert error. My command is
python train.py --c config/usb_cv/freematch/freematch_cifar100_400_0.yaml
![image](https://private-user-images.githubusercontent.com/36142776/284971709-30335d7d-0085-479a-b7a6-38fb7faa75c4.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.5_41Ax7FnCmFPbXfUX2w0NMthbTjHEHubB7CUbA96xc)
github-actions commented
This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.
github-actions commented
This issue was closed because it has been stalled for 5 days with no activity.