Why does the model have 32 quantizers? (and not 8, as mentioned in the paper)
· Sign up or log in to comment