CK framework documentation¶
Collective Knowledge framework (CK) helps to organize any software project as a database of reusable components (algorithms, datasets, models, frameworks, scripts, experimental results, papers, etc) with common automation actions and extensible meta descriptions based on FAIR principles (findability, accessibility, interoperability, and reusability).
The ultimate goal is to help everyone share, reuse, and extend their knowledge in the form of reusable artifacts and portable workflows with a common API, CLI, and JSON meta description.
See how CK helps to support collaborative and reproducible AI, ML, and systems R&D in some real-world use cases from Arm, General Motors, IBM, MLPerf(tm), the Raspberry Pi foundation, and ACM: https://cKnowledge.org/partners.
Introduction
Getting Started
- Prerequisites
- CK installation
- Docker
- Virtual CK environments with templates
- Customization
- Trying CK
- How CK enables portable and customizable workflows
- CK installation
- Pull CK repositories with the universal program workflow
- Manage CK entries
- Invoke CK automation actions
- Install missing packages
- Participate in crowd-tuning
- Use CK python API
- Try the CK MLPerf™ workflow
- Further information
- Contact the CK community
- The most common usage
- Initialize a new CK repository in the current directory (can be existing Git repo)
- Add dependency on other repositories to reuse automation actions and components
- Add a new program workflow
- Update software dependencies
- Reuse or add basic datasets
- Add new CK software detection plugins
- Add new CK packages
- Pack CK repository
- Prepare CK repository for Digital Libraries
- Prepare a Docker container with CK workflows
- Create more complex workflows
- Generate reproducible and interactive articles
- Publish CK repositories, workflows, and components
- Contact the CK community