topic Re: Managing the permissions using MLFlow APIs in Data Engineering

Managing the permissions using MLFlow APIs

SagarK1 — Thu, 21 Oct 2021 09:41:08 GMT

Hello All,

I am trying to manage the permissions on the experiments using the MLFLow API. Do we have any MLFlow API which helps to manage the permissions of Can Read ,Can Edit , Can Manage.

Example :

I create the model using MLFlow APIs and through my code I should be able to enforce the permission saying what permission can be given for this model.E.g. Can Read , Can Edit , Can Manage.

Thanks

Re: Managing the permissions using MLFlow APIs

Prabakar — Thu, 21 Oct 2021 19:48:20 GMT

Hi @Sagar Kant Please refer to the below document that provides a detailed explanation about the MLflow permissions API.

MLflow experiment permissions — Manage which users can read, edit, or manage MLflow experiments.

https://docs.databricks.com/dev-tools/api/latest/permissions.html#tag/MLflow-experiment-permissions

MLflow registered model permissions — Manage which users can read, edit, or manage MLflow registered models.

https://docs.databricks.com/dev-tools/api/latest/permissions.html#tag/MLflow-registered-model-permissions

Re: Managing the permissions using MLFlow APIs

Prabakar — Sun, 28 Nov 2021 22:04:20 GMT

Hi @Sagar Kant , Could you please let me know if the answer I provided earlier resolved your query? If so would you be happy to mark the answer as best so that others can quickly find the solution in the future?

Re: Managing the permissions using MLFlow APIs

sean_owen — Fri, 17 Dec 2021 02:33:13 GMT

I don't believe you can set MLflow permissions via the API. Permissions are a function of Databricks rather than MLflow itself.

Re: Managing the permissions using MLFlow APIs

jsan — Wed, 01 Mar 2023 12:35:00 GMT

Hey folks, did we get any workaround for this or what @Sean Owen said is true ?