Prune head_size
last layers of a tabnet network in order to
use the pruned module as a sequential embedding module.
nn_prune_head.tabnet_fit(x, head_size)
nn_prune_head.tabnet_pretrain(x, head_size)
nn_network to prune
number of nn_layers to prune, should be less than 2
a tabnet network with the top nn_layer removed