Artificial intelligenceDeep learningMachine learning
by CODAIT Team Published March 28, 2019
The IBM Model Asset eXchange launched in March 2018 with the goal of enabling developers and data scientists to more easily discover and use state-of-the-art, free and open source deep-learning models in their applications. This is achieved through careful vetting of code and licensing details, together with the creation of standardized microservices that are deployable in a matter of seconds with Docker or Kubernetes (see this blog post for a deep dive into the process behind wrapping models for the Model Asset eXchange).
In the past year, the exchange has grown to almost 30 models across a variety of common problem domains ranging from image recognition to natural language processing to time-series analysis. Together with the models themselves, we have focused on creating rich content to help you use these models, whether you’re just getting started with deep-learning capabilities in your applications or are already experienced. This content now includes:
In addition, because everything to do with the Model Asset eXchange is open, we’ve published our base Docker image, Model Asset eXchange starter project, and Model Asset eXchange Framework repos on GitHub for the wider community to use, comment on, and improve upon.
It’s been a busy year!
Now, just as a child often begins to walk at this age and explore a wider world, the Model Asset eXchange is widening its horizons. Together with a batch of new models, the CODAIT team has been hard at work enabling new deployment mechanisms for the exchange at the edge via the new module for Node-RED, which allows you to use Model Asset eXchange model microservices in your Node-RED flows by simply dragging and dropping – no code required! To illustrate this functionality, we have created a detailed tutorial.
We’re always working on adding new and useful models to the exchange. As mentioned, we’re pleased to announce a new batch of models encompassing audio, natural language processing, and image recognition:
In 2019, we’re excited to continue our mission to bring state-of-the-art deep-learning models to the developer community in ways that make this technology easier to deploy and consume. At the same time, for data scientists, we’re working on making these models easier to customize and apply to your own data in a standardized and frictionless manner. We’re also always thinking about ways to further standardize and modularize the way Model Asset eXchange model microservices are put together, in order to make it easier to deploy and consume these models across the platforms most used by developers — edge and IoT devices, the browser, mobile, cloud services, or your own Kubernetes cluster. Finally, the team will be out in the community, physically at events and virtually online, talking about and demonstrating the work we’re doing with the Model Asset eXchange.
Visit the IBM Model Asset eXchange and check out all the latest models and enhancements. We trust you will find the right solution for your AI development use cases — and if you don’t, let us know what you need and let’s work together to bring it to you and other developers! As always, we welcome your comments and suggestions that help us improve and better serve the ML/DL community. Reach out on GitHub or join the Model Asset eXchange Slack channel.
The Model Asset Exchange is place for developers to find and use free and open source deep learning models. Complete…
Artificial intelligenceData science+
Artificial intelligenceMAX - Model Asset eXchange
Learn how MAX is a place for developers to find and use free, open source, state-of-the-art deep learning models for…
Back to top