feat: add model#load_model method

[parano]

What does this PR address?

A slightly nicer way for loading model instances without thinking about which module to use - which is common when building custom runners or testing saved models. Here's an example of usage without vs. with this method added:

Let's say we saved a CLIP model from the transformer:

from transformers import CLIPModel, CLIPProcessor

model = CLIPModel.from_pretrained("openai/clip-vit-base-patch32")
processor = CLIPProcessor.from_pretrained("openai/clip-vit-base-patch32")

bentoml.transformers.save_model('CLIP:clip-vit-base-patch32', model, custom_objects={'processor': processor})

Without this method, loading the model and processor looks like this:

MODEL_TAG = 'clip:clip-vit-base-patch32'
model = bentoml.transformers.load_model(MODEL_TAG)
model_obj = bentoml.models.get(MODEL_TAG)
processor = model_obj.custom_objects['processor']

After this PR:

model_obj = bentoml.models.get('clip:clip-vit-base-patch32')
model = model_obj.load_model()
processor = model_obj.custom_objects['processor']

Fixes #(issue)

Before submitting:

• Does the Pull Request follow Conventional Commits specification naming? Here are GitHub's
  guide on how to create a pull request.
• Does the code follow BentoML's code style, both make format and make lint script have passed (instructions)?
• Did you read through contribution guidelines and follow development guidelines?
• Did your changes require updates to the documentation? Have you updated
  those accordingly? Here are documentation guidelines and tips on writting docs.
• Did you write tests to cover your changes?

[codecov]

Codecov Report

Merging #3780 (efc501e) into main (0107ead) will not change coverage.
The diff coverage is 0.00%.

@@          Coverage Diff          @@
##            main   #3780   +/-   ##
=====================================
  Coverage   0.00%   0.00%           
=====================================
  Files        154     154           
  Lines      12613   12619    +6     
=====================================
- Misses     12613   12619    +6     

Impacted Files	Coverage Δ
src/bentoml/_internal/models/model.py	0.00% <0.00%> (ø)

... and 1 file with indirect coverage changes

[aarnphm]

@parano I added a quick tests for testing this behaviour in transformers tests.

You can run it with pytest tests/integration/frameworks/test_frameworks.py tests/integration/frameworks/test_transformers_unit.py --framework transformers --capture=tee-sys