Can this detection library be used for training?

I have pip this package, And the effect is very good.

But I want to know if this can be used for training?

For example, I want to predict the landmarks of the generated image and the input image, and calculate the landmark loss to train the network.