LOS ANGELES—As this week’s Open Network Summit gets underway, The Linux Foundation has debuted the LF Deep Learning Foundation, an umbrella organization focused on driving open source innovation in artificial intelligence, machine learning and deep learning.
The goal of the LF Deep Learning Foundation is to make these new technologies available to developers and data scientists.
Founding members of LF Deep Learning include Amdocs, AT&T, B.Yond, Baidu, Huawei, Nokia, Tech Mahindra, Tencent, Univa, and ZTE. Through the LF Deep Learning Foundation, members are working to create a neutral space where makers and sustainers of tools and infrastructure can interact and harmonize their efforts and accelerate the broad adoption of deep learning technologies.
In tandem with the launch of LF Deep Learning, The Linux Foundation also debuted the Acumos AI Project, a platform that will drive the development, discovery, and sharing of AI models and AI workflows. AT&T and Tech Mahindra contributed the initial code for the Acumos AI Project.
These two companies are well positioned to be larger contributors to the AI projects. It was just last fall that AT&T partnered with Tech Mahindra to create its artificial intelligence (AI) platform Acumos, with plans to bring AI solutions into the broader open source community.
Hosted by The Linux Foundation, Acumos will enable developers to build, share and deploy AI applications, according to AT&T. The platform will provide a marketplace for accessing, using and enhancing those applications.
There are five components to any deep learning system: the models and algorithms, the data set, the running environment, the finished services and use cases.
“What Acumos is aiming to do and LF is aiming to do in terms of the deep learning project is to bring this technology to mass market, meaning create a marketplace and commoditize that,” said Arpit Joshipura, general manager of networking and orchestration for The Linux Foundation, in an interview with FierceTelecom.
AT&T could create a template to direct drones that inspect a wireless tower to send data about rust or other corrosion into ONAP. A DCA module implements could then implement a policy where it creates a trouble ticket based on some learning they have done, for example.
But this is not just relegated to AT&T alone.
“The template AT&T not proprietary to just AT&T,” Joshipura said. “They have done the work to integrate it with a data license, deep learning tools and that could also be used by China Mobile, for example.”
Joshipura added “what Acumos brings is a marketplace to share templates within verticals—networking verticals, security verticals, financial and others.”
A key goal of the project is to make the AI accessible to all with methods to create, share, discover and apply machine learning, deep learning and analytics models. The Linux Foundation will host the Acumos AI platform and the Acumos Marketplace moving forward with the goal of nurturing an active, large ecosystem around the project to sustain it over the long term. Code is already available for download.
Besides the Acumos AI Project, LF Deep Learning anticipates future project contributions from Baidu and Tencent, among others. Baidu’s EDL project enhances Kubernetes with the feature of elastic scheduling and uses PaddlePaddle’s fault-tolerable feature to significantly improve the overall utilization of Kubernetes clusters.
Tencent’s Angel project, a high-performance distributed machine-learning platform jointly developed by Tencent and Peking University, is tuned for big data/models. Linux Foundation said it can support over 1 billion parameters.