Is the pre-trained model uploaded the same that achieved the results in the paper?

[fjzs]

Thanks

[jiaxue-ai]

I've adapted the code to PyTorch, which may introduce some training discrepancies. But you should still achieve comparable classification accuracy.