GitHub - narelabs/amlb: A decoupled context-compression co-processor utilizing gated cross-attention for linear-time, O(1)-memory sequence scaling in frozen large language models. (159 символов)

Name		Name	Last commit message	Last commit date
.gitattributes		.gitattributes

About

A decoupled context-compression co-processor utilizing gated cross-attention for linear-time, O(1)-memory sequence scaling in frozen large language models. (159 символов)

Activity

Custom properties

0 stars

0 watching

0 forks

Report repository

Navigation Menu

About

Releases

Packages

Contributors

Latest commit

History

Folders and files

About

Stars

Watchers

Forks

Releases

Packages

Contributors