How calculate loss.backward() in MLM by PyTorch

Models
2

Hey @MahdiA what you have looks pretty close - I think you just need loss.backward() instead of loss.backwards()

Does that solve your issue?

1 Like