medium10 pts · first answer countsinference-optimisation
Smaller model with similar quality
A large model meets quality targets but is too slow for a 50 ms latency budget. Which technique trains a smaller model to imitate it?
Discussion 0
Join free to post in the discussion.