Custom flags required with docker run, is this possible with online endpoint

adm.t1.mowen 0 Reputation points
2025-04-09T15:12:20.09+00:00

My custom Docker environment breaks unless I can provide the following flags with the docker run command: --gpus all --ipc=host --ulimit memlock=-1 --ulimit stack=67108864. This has led to me being unable to deploy endpoints either locally, or online, since presumably the docker run command is being handled in the backend by azure ml. So my question is, can I somehow include custom docker run commands on deployment, or is there a workaround which would allow me to use this environment with an endpoint deployment?

Thanks in advance, and apologies if I am asking a naïve question, I am quite new to this. I could not find anything in the documentation that seemed applicable.

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
3,337 questions
{count} votes

1 answer

Sort by: Most helpful
  1. Saideep Anchuri 9,500 Reputation points Moderator
    2025-04-10T17:22:42.5133333+00:00

    Hi adm.t1.mowen

    You have to use build parameter in environment class to provide dockerfile path with the required parameters and then use this environment in the deployment of online endpoint.

    Kindly refer below links: Environmentclass

    Deploya model in a custom container to an online endpoint

    Thank You.

    0 comments No comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.