Acumos, an LF AI Foundation Graduated Project, has announced their fourth software release, codenamed Demeter. We’re thrilled to see another great release from the community!
This new version introduces a fully cloud-enabled platform which is quick and easy to deploy and features elastic performance based upon load, making it suitable for use by small teams or large companies. Demeter also introduces bidirectional communications on the federation link, allowing developers to securely receive information about model performance and provide users with automatic and timely updates.
Acumos is a platform and open source framework that makes it easy to build, share, and deploy AI apps. Acumos standardizes the infrastructure stack and components required to run an out-of-the-box general AI environment. This frees data scientists and model trainers to focus on their core competencies and accelerates innovation.
Major highlights of the Demeter release include:
- Containerized platform deployment incorporating cloud native functions, horizontal scaling, and implementation flexibility.
- CLI message response with the Acumos Docker model.
- Support for Pre-dockerized and Dockerized model URI with protobuf file to render models usable in Design studio.
- Activity tracking and reporting – License usage manager (LUM) maintains logs of model usage.
- Integration of License module with Portal UI.
- Bidirectional communication over the federation link between subscriber and supplier instances to support ML life cycle management and continuous learning.
- Predictor Manager
- The Predictor Manager manages the model deployment, visualization of deployment metadata and association to a project.
- Data source
- The Data Source feature allows users to create and associate project data with a model to create, update, and delete data sets used for training, validation and testing.
To learn more about the Acumos Demeter release, check out the full release notes. Want to get involved with Acumos? Be sure to join the Acumos-Announce and Acumos Technical-Discuss mailing lists to join the community and stay connected on the latest updates.
Congratulations to the Acumos team and we look forward to continued growth and success as part of the LF AI Foundation! To learn about hosting an open source project with us, visit the LF AI Foundation website.
Mazin Gilbert, VP of Technology and Innovation, AT&T, said: “AT&T is proud to be a founding member of the Acumos platform, which has radically improved over the past 2 years. As AI begins to play a central role in 5G, openness and industry collaboration are proving more valuable than ever. Acumos has brought together some of the brightest minds in the industry, and we’re consistently encouraged by the progress.”
Sachin Desai, LF AI Board Member and Ericsson VP of Global AI Accelerator, said: “In the Acumos Demeter release, Ericsson has contributed in areas such as bidirectional communications and enhanced licensing functionalities which enable continuous training, transfer learning and its commercialization. This new release will enable a broader range of applications and collaborations among Acumos ecosystem partners and support more complete life cycle management of AI in telecommunication. We are excited to continue providing Ericsson’s technology leadership related to AI/ML in telecommunication domain through our active engagement in the open source community.”
Jonne Soininen, Head of Open Source Initiatives at Nokia, said: “Acumos with the Demeter Release offers a complete experience for ML modelers and model consumers beyond the marketplace. As a founding member of the LF AI, Nokia is excited about the possibilities of Acumos in powering ML model marketplaces around network automation. We are always impressed with the great results of the Acumos community and the Demeter release is no different.”
Emmanuel Lugagne Delpon, SVP Orange Labs Networks and Group CTO, said: “The Demeter Release of Acumos is a new step towards a seamless way to produce and deploy AI models in various environments such as 5G and ONAP network automation: this will be eased thanks to the Cloud Enablement and Predictor Manager features. The addition of Data Source – a keystone of each AI project – is also a crucial feature for companies who want to manage all of their AI models in a centralized Marketplace catalogue (from the model training to the deployment). The development of the Demeter release during the unique situation the world is facing today is really impressive: it demonstrates the strength of open source communities combined with individual commitments of each contributor. To favor cross cooperation between communities and to meet its operational objectives, Orange is also active in the integration of this new release with ONAP in order to test 5G use cases.”
Sachin Saraf, SVP and VBU Head, CME Business, Tech Mahindra, said: “Tech Mahindra is excited to see Acumos AI reaching its fourth major milestone in a short span of two years with the release of Demeter. Seamless hosting of the Acumos AI platform has been energized by the Cloud Enablement feature. Predictor management, as well as the data source features of ML workbench, which allows users to manage and associate the models to the project and data. As part of our TechMNxt charter, the latest Acumos AI release fosters co-innovation and co-creation in the field of AI and ML and is a mainstay of our key focus areas. We will continue to work towards the enrichment of Acumos AI platform. Our enterprise grade solutions are anchoring a key role in accelerating the adoption of Linux Foundation Artificial Intelligence open source projects.”
Bingtao Han, Chief System Architecture Expert, ZTE, said: “Acumos Demeter release marks a significant milestone for Acumos development with enhanced support for bringing ML models into production, like much more flexible cloud enablement and deployment management of ML models. ZTE believes that Acumos has great potential to be a key enabler of network intelligence in 5G evolution. We’ll continue to support Acumos and embrace the opportunities in open source practices.”
Acumos Key Links
LF AI Resources