6 ms·

Couldn't they just grab and run an open weight model to save on API tokens?

• onel 6d ago
  You get better performance if you also finetune it for your task