Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

20 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Transformer Decoder

Objective

The existing transformer libraries doesn't have KV cache and wasn't flexible to use as per the need. This will solve the problems those implementations have. This is built entirely on C++ with the use of Libtorch library (Pytorch with C++ API) faster and flexible

Working

The Transfomer takes a struct "TransformerConfig" which includes the attributes of the individual components of entire model.

About

C++ implementation of flexible, customizable and efficient transfomer-decoder architecture. built entirely using C++ and Libtorch ( C++ implementation of Pytorch)

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages