注册并分享邀请链接,可获得视频播放与邀请奖励。

Philip Kiely
@philipkiely
Author of Inference Engineering | Building @baseten | Not an LLM (yet)
加入 December 2018
988 正在关注    16.6K 粉丝
Scaling past a trillion parameters requires inference engineers to solve 3 issues: - Width: Routing between hundreds of sparse experts - Depth: Mixing global and local linear attention - Length: Processing 1M-token sequences I asked Ali and Bryce to draw it up. They delivered.
显示更多
0
41
2.1K
211
转发到社区