Sign up to take part
Registered users can ask their own questions, contribute to discussions, and be part of the Community!
The downloaded API package version is used to directly upload the version to a deployer (as shown below). You can also use it to share with team members and troubleshoot.
Technically yes but why would you reinvent the wheel? At the end of the day the package it's Python code so you could write a service that can run it as the API node does. But your time will be best used developing machine learning models rather than building infrastructure you already have available.
Hi, Turribeach, thanks for reply. The dataiku I using is hosted on my company cloud environment, the authorization process is very complicated and time consuming. Another reason is that my algorithm is based on customized code, I couldn't download the trained model as python or MLFlow code.