Accessing inference endpoint environment / config from within the inference code is undocumented

[ram-nadella]

What did you find confusing? Please describe.
I was looking for a way to get information about the inference endpoint from the inference endpoint code at runtime (eg. within a Flask request handler). Similar to how training and processing have access to JSON config in a file under the /opt/ml/ structure or an env var.

Describe how documentation can be improved
Need docs for the right /opt/ml/ config file that would provide info about the current inference endpoint (the name would be a good starting point as the rest of the info about model etc. can be pulled from this using the SDK)

Additional context
Example of how this info is accessible for training:

SM_TRAINING_ENV has a lot of info about the job itself.
https://docs.aws.amazon.com/sagemaker/latest/dg/docker-container-environmental-variables.html

Is there a similar file/env var for inference?

FYI, we are using a custom container for inference.

[laurenyu]

@ram-nadella sorry for the delayed response here

is your custom image based on one of the prebuilt SageMaker images or completely built from scratch?

[ram-nadella]

Hi @laurenyu it's completely from scratch.

(we're operating in a python environment, using a base of continuumio/miniconda3:4.6.14)

[laurenyu]

Thanks for clarifying! I've passed this along to the SageMaker Hosting team to see if they can help.