[ Back to index ]

Collective Knowledge v1 and v2 (CK)

The open-source Collective Knowledge Technology v1 and v2 (CK) was originally developed by Grigori Fursin with great contributions from the community.

CK was based on Grigori's practical experience helping the community reproduce many research projects and validate them in the real world across diverse and continuously changing models, data sets, software and hardware since 2013.

CK development was also sponsored by the cTuning foundation, HiPEAC and OctoML.

The archive of this discontinued framework is available in the /ck directory.

MLCommons Collective Mind (CM)

Grigori donated the CK technology to MLCommons in 2022 to benefit everyone, prototyped a new version of CK called Collective Mind (CM) with portable and reusable automation recipes for universal benchmarking and optimization of applications and systems (CM4MLOps scripts and workflows), and helped establish the MLCommons Task Force on Automation and Reproducibility co-led with Arjun Suresh.

We continue extending CM to support different MLCommons projects to modularize and unify benchmarking of ML/AI systems as a collaborative engineering effort based on user feedback.

You can learn more about the CM concept and motivation from the keynote at ACM REP'23 and this white paper.