It's a painful, ad-hoc, time consuming and expensive process to develop novel applications based on emerging technologies (ML, AI, HPC, quantum, IoT) and deploy them in production due to continuously evolving software, hardware, models, data sets and research techniques.
We are developing the open-source Collective Knowledge framework (CK) to bring DevOps practices to research and help researchers and practitioners design, benchmark, optimize and validate such application.
CK can help to convert ad-hoc artifacts and workflows into reusable, portable, customizable and non-virtualized CK components with a unified CLI, automation actions, Python API and JSON meta description. Our idea is to gradually abstract all existing artifacts (software, hardware, models, data sets, results) and use the DevOps methodology to connect such components together into functional CK solutions. Such solutions can automatically adapt to evolving models, data sets and bare-metal platforms with the help of customizable program workflows, a list of all dependencies (models, data sets, frameworks), and a portable meta package manager.
More resources:
Reaching this goal will help us to support the community initiatives at AI, ML and systems conferences to implement portable workflows for the state-of-the-art research techniques from published papers and reproduce their results.
0% towards $1,000 per month goal
Be the first to sponsor this goal!
$1,200 a month
SelectReaching this goal will help us to support the community initiatives at ML and systems conferences to implement portable workflows with reusable artifacts for the state-of-the-art research papers and reproduce their results.