Bug: image_name is missing from command arguments

[aShevc]

Summary / Background

Currently unable to set image_name when declaring sagemaker deployment, hence can not use prebuit images. The command attempted:

mlem declare deployment sagemaker app.mlem \
    --env env.mlem \
    --method predict \
    --use_prebuilt True \
    --instance_type ml.t2.medium \
    --env.ecr_repository 763104351884.dkr.ecr.us-east-2.amazonaws.com \
    --image_name pytorch-inference \
    --image_tag 1.12.1-cpu-py38

This returns:

No such option: --image_name (Possible options: --image_tag, --model_name)

When image_name is removed , but use_prebuilt is left intact, runnonig the command results in

image_name not specified

This blocks an ability to run prebuilt images
`