https://github.com/matrix97317/oneneuralnetwork
This is a cross-chip platform collection of operators and a unified neural network library.
https://github.com/matrix97317/oneneuralnetwork
ai-compiler compiler cuda-kernels deep-neural-networks deeplearning-framework nerual-networks pytorch
Last synced: 4 months ago
JSON representation
This is a cross-chip platform collection of operators and a unified neural network library.
- Host: GitHub
- URL: https://github.com/matrix97317/oneneuralnetwork
- Owner: matrix97317
- License: apache-2.0
- Created: 2023-11-01T06:55:00.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2023-11-03T11:57:07.000Z (about 2 years ago)
- Last Synced: 2024-05-02T05:42:48.947Z (over 1 year ago)
- Topics: ai-compiler, compiler, cuda-kernels, deep-neural-networks, deeplearning-framework, nerual-networks, pytorch
- Language: Python
- Homepage:
- Size: 103 KB
- Stars: 3
- Watchers: 2
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# One Neural Network (ONN)
This is a cross-chip platform collection of operators and a unified neural network library.
We hope to use this lib to run [OpenMMLab/Detectron/DeepSpeed/Magatron-LM/...] on different chip platforms.
To achieve this goal, we utilized Deep Learning Compiler or DSL (such as,TVM,Triton,... etc) as the backend.
At the same time, we abstracted the operators from some high-level applications.
### Developer Quick Start
If you want to be a developer, you can read this [developer_manual](./md_doc/developer_manual.md)
### User Quick Start
As a user, you can to complete cross chip platform training and inference based on ONN. You can read this [user_manual](./md_doc/user_manual.md)
### Design Manual
You can read this [Design Manual](./md_doc/design_manual.md) for more detail.
### Acknowledgements
- Project Scaffold，From：https://github.com/serious-scaffold/serious-scaffold-python