As described in the MindSpore governance, Special Interest Groups (SIGs) are persistent groups responsible for specific parts of the project. SIGs have open and transparent proceedings to develop goals and implement code contributions. SIGs are also responsible for ongoing maintenance of the code in their areas.
New SIGs are created when there is sufficient interest in a topic area and someone volunteers to be the lead for the group and submits a proposal to the steering committee. The chair facilitates the discussion and helps synthesize proposals and decisions.
If you are interested in participating, please join the discussion in the respective list. Details about any upcoming meetings will also be shared in the mailing list. SIG artifacts can be found in the current repository.
| SIG name | Responsibilities | SIG Leads |
|---|---|---|
| FrontEnd | This SIG is responsible for the development of MindSpore front-end expression. | @kingxian |
| Compiler | This SIG is responsible for the development of MindSpore high level graph compilation. | @zh_qh |
| Executor | This SIG is responsible for the development of MindSpore back-end support for pipeline. | @kisnwang |
| ModelZoo | This SIG is responsible for the development of MindSpore modelzoo and additional ops. | @yingjy |
| Data | This SIG is responsible for the development of MindSpore data processing and data format transformation. | @liucunwei |
| GraphEngine | This SIG is responsible for the development of MindSpore graph engine for Ascend AI processor. | @youui |
| Visualization | This SIG is responsible for the development of MindSpore visualization tools. | @gaocongli_hw |
| Security | This SIG is responsible for the development of MindSpore security related tools. | @randywangze |
| AKG | This SIG is responsible for the development of MindSpore auto kernel generator. | @anyrenwei |
| MSLITE | This SIG is responsible for the development of MindSpore lite. | @zhaizhiqiang |
| MDP | This SIG is responsible for the development of MindSpore programming library for Bayesian deep learning. | @jianfeichen([email protected]) |
| Parallel | This SIG is responsible for the development of MindSpore's functionality of automatically finding the efficient parallel strategy for DNN training and inference. | @dr-orange([email protected]) |
| AdaptiveTraining | This SIG is to develop an adaptive distributed training system that can train the neural networks in elastic clusters without affecting the convergence. | @luomai-edin([email protected]) |
| Serving | This SIG is responsible for the development of MindSpore Serving module. | @xu-yfei |
| DevelopereXperience | This SIG is responsible for improving the experience of those who upstream contribute or develop applications for MindSpore community. | @jiancao81([email protected]) @clement_li |