feature: add support for online ML models from River library in BentoML

[jsamantaucd]

Feature request
We have been developing and experimenting river (https://riverml.xyz/latest/api/overview/) for training and inferencing time series and continuous machine learning tasks and would like to add support for river in BentoML.

We have also created a custom python module to add river models using MLflow which can be saved in the BentoML model's repository and later can be retrieved from the BentoML for inferencing.

Motivation
River offers a diverse selection of ML models for online learning that are much more efficient than the sklearn models, and integrating this MLflow would create a comprehensive ML solution.

Others:
This feature was tested on tag v1.0.20

[jsamantaucd]

This PR is related to the feature request: #4640

[frostming]

Hi @jsamantaucd We are preparing to restructure the examples and do not accept new framework-specific examples.

You can make it in a standalone repository and we can link it in our README, thanks.

[jianshen92]

@jsamantaucd Here is an example on how we showcase examples, https://github.com/bentoml/BentoVoiceAgent

[jsamantaucd]

Hi @frostming and @jianshen92, thank you for your valuable feedback, I have created a standalone repository as suggested: https://github.com/jsamantaucd/BentoRiverModel, please add it in your README. Thanks!