ALMA and eProsima are happy to announce the first release of the AML Integrating Platform (AML-IP), the platform that allows for easy Algebraic Machine Learning (AML) node communication.
Developed under the umbrella of the ALMA project, AML-IP is a communications framework designed to allow non-experts users to create and manage a cluster of AML and ML nodes to exploit the distributed and concurrent learning capabilities of AML and ML. The new AML-IP v0.1.0 is set to redefine the landscape of data exchange among AML nodes and other distributed ML architectures, making it easier than ever to harness the potential of distributed and concurrent learning.
Empowering AML-IP Nodes and Deployment Scenarios
In order to enhance the AML-IP capabilities, new AML-IP Nodes have been included.
AML-IP v0.1.0 enables a range of deployment scenarios, including Monitor State, Workload Distribution, Collaborative Learning, and Distributed Inference. The inclusion of nodes like Status, Main, Computing, Edge, Inference, Agent, Client, Server, Repeater, Model Manager Receiver, and Model Manager Sender widens the horizons of potential use cases.
Streamlining User Experience
AML-IP’s first release introduces a slew of features aimed at streamlining user experience:
- C++ and Python APIs
- Implementation of DDS entities
- Implementation of Multiservice and Asynchronous Multiservice protocols
- Internal package 'amlip_demo_nodes' for demos
- Dockerfile for creating AML-IP docker image
- Custom RPC communication over DDS
Get Started with AML-IP Demos
This release includes the following demos with different scenarios that focus not only on AML but also on other ML technologies such as DNN (Deep Neural Networks) for Object Detection built with Tensorflow framework to accomplish a distributed ML environment:
- Collaborative Learning
- TensorFlow Inference
- TensorFlow Inference using ROSbot 2R
- Workload Distribution
Find all the information on the AML-IP v0.1.0 in the release notes.
MORE INFORMATION ABOUT ALMA:
For any questions please contact
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 952091.