Lingsoft arranged a Docker and API workshop for beginners as part of our "Microservices at Your Service" project. Filip Ginter and Juhani Luotolahti from Turku NLP group assisted Lingsoft's Sebastian Andersson with the workshop presentations. The workshop is targeted at people interested in speech and language technology development who are beginners at REST API and Docker. The whole workshop was recorded and can be watched here. The presentations contain hands-on examples with code that can also be downloaded and tried.
Docker offers a convenient way of putting software tools and models into a package that makes it very easy for other software developers and data scientists to use those tools and models. According to Wikipedia: “[Docker is] a set of platform as a service (PaaS) products that use OS-level virtualization to deliver software in packages [...]”. The created Docker package is called an "image". This image is a file that can be copied and shared with just your colleagues or it can be registered with e.g. DockerHub and shared with the world!
By adding a web service API to your software and putting it on a Docker one can arrange many different types of tools and models into a microservice architecture similar to how the European Language Grid (ELG) and many other Saas/PaaS are organised, including Lingsoft's own SaaS. The main benefit is that it is very easy to replace outdated tools/models with state-of-the-art from e.g. the great speech and language technology open source communities on GitHub, ELG, Hugginface, etc. This is very important in the fast moving field of artificial intelligence, including speech and language technology.