Viewing a single comment thread. View all comments

zaptrem t1_ixbniz5 wrote

Could these advancements in quantization be used to train in int8 as well?

3