search

mlcommons / ck

Collective Mind (CM) is a small, modular, cross-platform and decentralized workflow automation framework with a human-friendly interface and reusable automation recipes to make it easier to build, run, benchmark and optimize AI, ML and other applications and systems across diverse and continuously changing models, data, software and hardware
https://access.cKnowledge.org/challenges
Apache License 2.0
597 stars 110 forks source link

Suggestions from the CK survey #132

Closed gfursin closed 2 years ago

gfursin commented 3 years ago

Here are some suggestions from our CK survey: 1) moving CK to a foundation such as LF AI or MLCommons to provide better governance and community support 2) changing license to Apache 2.0 3) dropping support for Python 2 and keeping support for Python 3.6+ 4) rewriting CK to be more pythonic (on-going project in the CK incubator) 5) including CK modules to the CK core to ensure the stability of shared workflows 6) making it easier to add ML workflows (providing a GUI at cKnowledge.io) 7) continue automating MLPerf: https://github.com/ctuning/ck/blob/master/docs/mlperf-automation/README.md 8) adding more CK components and workflows for ML (BERT, DLRM): on-going 9) collecting all CK components in one GitHub repository 10) providing better intergration with existing tools and projects including Kuberflow and PwC 11) adding super simple CK tutorials

gfursin commented 3 years ago

I addressed suggestions #5 and #9 in the latest CK release V1.50.0:

gfursin commented 3 years ago

Fixed broken release: https://github.com/ctuning/ck/releases/tag/V1.55.0

gfursin commented 3 years ago

Some notes about the new development methodology: https://groups.google.com/g/collective-knowledge/c/0cGOPNM1Uvw

gfursin commented 3 years ago

Forgot to mention CLA: https://github.com/ctuning/ck/issues/135

gfursin commented 3 years ago

We relicensed to Apache 2.0

gfursin commented 3 years ago

We work with the community to revise, test, reproduce and stabilize MLPerf workflows: https://github.com/ctuning/ck/blob/master/docs/mlperf-automation/README.md Don't hesitate to join this effort!

gfursin commented 2 years ago

Many of these suggestions were implemented in the latest version of the CK prototype. CK is stable and has been successfully used by different companies and organizations to automate MLPerf inference submissions. That is why we have decided not to touch it anymore expect bug fixes in the CK kernel and existing MLOps components. Instead, we are developing a new CK2 framework based on user feedback.

I close this issue.