FP8 versus INT8 for efficient deep learning inference

1 pointsposted 11 hours ago
by lnyan
(arxiv.org)

No comments yet