TL;DR
- takes arbitrary assets (Jupyter notebooks, python scripts, R scripts) as input
- automatically creates container images and pushes to container registries
- automatically installs all required dependencies into the container image
- creates KubeFlow Pipeline components (target workflow execution engines are pluggable)
- creates Kubernetes job configs for execution on Kubernetes/Openshift clusters
- can be triggered from CICD pipelines
C3 (CLAIMED Component Compiler) is the central project of the CLAIMED framework. It automates the transformation of arbitrary code assets — such as Jupyter notebooks, Python scripts, or R scripts — into fully portable, executable AI components.
While the component library is now maintained primarily as an example repository, C3 is where active development and innovation take place. The most utilized and powerful feature of C3 is grid compute parallelization, enabling distributed execution of AI workloads across heterogeneous compute environments.
The Machine Learning eXchange (MLX) is now fully integrated as the backend for C3’s grid computing system, responsible for tracking all assets, including:
-
data
-
models
-
jobs
-
and other related resources
This integration allows C3 to seamlessly manage asset lifecycle, provenance, and discovery within a unified infrastructure.
To learn more on how this library works in practice, please have a look at the following video
pip install claimedJust run the following command with your python script or notebook:
c3_create_operator "<your-operator-script>.py" --repository "<registry>/<namespace>"Your code needs to follow certain requirements which are explained in Getting Started.
c3_create_operator --helpWe welcome your questions, ideas, and feedback. Please create an issue or a discussion thread. Please see VULNERABILITIES.md for reporting vulnerabilities.
Interested in helping make CLAIMED better? We encourage you to take a look at our Contributing page.
CLAIMED is supported by the EU’s Horizon Europe program under Grant Agreement number 101131841 and also received funding from the Swiss State Secretariat for Education, Research and Innovation (SERI) and the UK Research and Innovation (UKRI).
This software is released under Apache License v2.0.