Comment on Honda: 2 years of ml vs 1 month of prompting - heres what we learnedComments−suddenlybananas9moI didn't read this very carefully so maybe I missed it, but I'm surprised they didn't try using a classifier on top of a BERT-style encoder model?
Comments
I didn't read this very carefully so maybe I missed it, but I'm surprised they didn't try using a classifier on top of a BERT-style encoder model?