Skip to content

Pre-made chains/templates for chaiNNer. Pytorch, NCNN, and ONNX included.

License

Notifications You must be signed in to change notification settings

Kim2091/chaiNNer-Templates

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

56 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

chaiNNer-Templates

These are my pre-made chains/templates for chaiNNer: https://github.com/chaiNNer-org/chaiNNer

You'll likely notice that there are 3 folders: NCNN, ONNX, and Pytorch. These all have different use-cases:

  • Pytorch - Best option for Nvidia GPUs, but can also work on CPUs. If you're on Linux and configured ROCM, this can be used for AMD as well
  • NCNN - Best option for AMD and Intel GPUs, but also supports Nvidia
  • ONNX - Works on Nvidia and CPU. This is usually slower, but is the only option for certain models

Each type of chain is in each folder. They're all identical but with each processing type replaced. Here's an example:

image

About

Pre-made chains/templates for chaiNNer. Pytorch, NCNN, and ONNX included.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published