The tug-of-war between cache and capacity: from MHA, MQA, GQA to MLA Posted by YuxiLiuWired on Monday February 03, 2025 @01:12AM from the yuxi-liu-wired.github.io dept. https://yuxi-liu-wired.github.io/docs/posts/2024-05-13-multi-latent-attention/ ( no comments | on HN )
There are no comments yet.