longlnOff
Results
2
issues of
longlnOff
How can I pass mask argument to use local attention instead of global attention in MHA module? Thank Sir!
Sir, I get this error when run your code: Found inputs tensors to be on [device(type='cuda', index=0), device(type='cuda', index=0), device(type='cuda', index=0), device(type='cuda', index=0)] while needed to be on a 'cuda'...