fp16 and fp32 doesn't change the generation time

by troilus - opened May 2, 2023

May 2, 2023

both torch.float16 and torch.float32 having the same generation speed (~7s per 20 steps in GPU)

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment