[FEATURE] Support register_model with a dict parameter model_config

[yuye-aws]

Is your feature request related to a problem?
The register_model method in ml_commons_client.py only accepts model_config_path as the input for model_config.

What solution would you like?
A clear and concise description of what you want to happen.

The register_model method can also take in a dict parameter such as model_config

What alternatives have you considered?
A clear and concise description of any alternative solutions or features you've considered.

Do you have any additional context?
Add any other context or screenshots about the feature request here.

[Arnav-Gr0ver]

Hello @yuye-aws, wanted to clarify something on this feature.

I checked the register model method and when the model is uploaded there is a check against the mandatory fields - a dict parameter such as model_config doesn't contain all the necessary parameters i.e. name, version, format in comparison to current implementation of a json file with all the metadata.

So, are you suggesting there should be an alternate parameter for a dict which contains all the mandatory parameters along with the model config dict (guessing this would be a nested dict)?

[dblock]

Catch All Triage - 1 2 3 4 5 6

[dhrubo-os]

@yuye-aws could you please clarify your question further?

[yuye-aws]

So, are you suggesting there should be an alternate parameter for a dict which contains all the mandatory parameters along with the model config dict (guessing this would be a nested dict)?

Exactly yes. When calling register_model method, not every user wishes to create a json file as the model config. It would be better if this method accepts a new dict parameter like model_config.