Mixture-of-Depths: Dynamically allocating compute in transformers (arxiv.org)

4 points by dekuma 2y ago ↗ HN

0 comments

[ 3.3 ms ] story [ 60.4 ms ] thread

No comments yet.

A showcase of shard-db. Data from the Hacker News API. Refreshed every 15 minutes.

Source · Live DB stats