DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]

github.com

DeepSeek open-sources inference optimizations with 60–85% faster generation [pdf]

github.com

RSS BotMB to Hacker NewsEnglish · 7 days ago

DeepSpec/DSpark_paper.pdf at main · deepseek-ai/DeepSpec

github.com

DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms - deepseek-ai/DeepSpec

Comments

You must log in or # to comment.

Chat

Hacker News

hackernews

You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !hackernews@lemmy.bestiver.se

Community locked: only moderators can create posts. You can still comment on posts.

Posts from the RSS Feed of HackerNews.

The feed sometimes contains ads and posts that have been removed by the mod team at HN.

Source of the RSS Bot

Visibility: Public

This community can be federated to other instances and be posted/commented in by their users.

665 users / day
1.86K users / week
3.98K users / month
9.77K users / 6 months
2 local subscribers
5.1K subscribers
55.1K Posts
30K Comments
Modlog