Accept container family as input for model creation and param validation

[VipulMascarenhas]

Description

This PR includes the changes for:

1. When user registers the model and inference container is not set, then set default as odsc-tgi-serving.
2. When deployment is created, accept container_family as input when user deploys unverified models.
3. When model params are being validated, accept container_family as input when user validates params for unverified models.

Unit Tests

> python -m pytest -q tests/unitary/with_extras/aqua/test_deployment*.py
==================================================== test session starts =====================================================
platform darwin -- Python 3.8.18, pytest-7.4.0, pluggy-1.0.0
rootdir: /Users/user/workspace/git/accelerated-data-science
configfile: pytest.ini
plugins: Faker-24.9.0, anyio-4.2.0
collected 29 items

tests/unitary/with_extras/aqua/test_deployment.py ....................                                                 [ 68%]
tests/unitary/with_extras/aqua/test_deployment_handler.py ..s......                                                    [100%]

=============================================== 28 passed, 1 skipped in 7.69s ================================================

[mayoor]

This should be based on the check if this model has text_generation_inference tag. If not, we should default to vllm

[VipulMascarenhas]

updated this to get the right image family based on the text_generation_inference tag. I thought all text-generation models will have the tag anyway, so didn't add an additional check here.

[VipulMascarenhas]

fyi - moved this from datasciencemodeldeployment.enums to here to avoid circular imports.

[mrDzurb]

NIT: shouldn't we log here the actual error, not the one that we generated?

[VipulMascarenhas]

not for this one. We want this to be a user friendly message when this entry is not available in custom metadata.