Skip to content

imxieyi/CoreML-MPS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CoreML-MPS

Introduction

Inspired by: A peek inside Core ML

This is a proof-of-concept project. It directly reads weights and bias from compiled CoreML v1 model and use them in MPSNN.

The demo is an implementation of waifu2x-ios.

Warning

This is generally a hack and is not documented by Apple in any forms. DO NOT try to use this in production environment since you will never know when Apple will change the compiled mlmodel format.

The maximum size of image supported is dependent on actual device. The GPU memory must be able to hold temporary texture of maximum number of feature channels. For details please refer to Metal Feature Set Tables.

Screenshot

demo

Image source: https://www.pixiv.net/member_illust.php?mode=medium&illust_id=45068168

About

Run compiled CoreML(v1) model using MPSNN

Topics

Resources

License

Stars

12 stars

Watchers

1 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages