Hi friends,
I’m working with some semantic data. I have ~2000 potential labels and looking for the best technique/model to autonomously predict the label. Does anyone recommend a model that would work best? Is fine-tuning ridiculous with that many possible labels?
Thanks!