[ot][spam][crazy] adapters for semibalanced trees?

[Undiscussed Horrific Abuse, One Victim of Many]

the final trained adapter is only a few megabytes large, despite the
model being several gigabytes.

a nagging part of me keeps considering the pretraining content of the
model. i'm not sure what t5 models are trained but, i imagine a
generative or masked model would have more general-purpose knowledge
and general language understanding, than something that was simply
trained on translation or summarization type tasks, where word-mapping
without comprehension could get you very far. i should probably look
this up at some point. but anyway i'm thinking of switching to xlnet.