Suggestion requested for use case

SK

Hi,

Can we create a generic framework(configuration based architecture) on dataiku for our machine learning models which share's similar objective with each other, what I mean is can we create a generic recipe which reads parameter values from other recipes and execute as per passed parameter.

Can you please share some Dataiku reference material using which such use case can be implemented.

Thanks!

Andrey

Hi @SK
,

Not sure I fully understand your question, but you can indeed create custom recipes in DSS with custom parameters and logic. The way to do it would be to create a plugin and implement the recipe in python. The logic may vary depending on the input parameters that are also customizable.

Please refer to these pages for more details:

https://doc.dataiku.com/dss/latest/plugins/reference/recipes.html

https://doc.dataiku.com/dss/latest/plugins/reference/other.html#custom-settings-ui

SK

Yeah thanks for information, I have created plugin's to cater client need