Unable to load mlflow model from /mnt/azureml/cr/j/54eb22ef76e94a02856ed4611dba04ae/cap/data-capability/wd/INPUT_model_input in current environment due to error: cannot import name 'logsumexp' from 'sklearn.utils.fixes' (/azureml-envs/responsibleai/lib/python3.8/site-packages/sklearn/utils/fixes.py)

### Operating System

Linux

### Version Information

This issue occurs in Azure ML Studio when trying to create an RAI dashboard.

ValueError: Unable to load mlflow model from /mnt/azureml/cr/j/54eb22ef76e94a02856ed4611dba04ae/cap/data-capability/wd/INPUT_model_input in current environment due to error:
cannot import name 'logsumexp' from 'sklearn.utils.fixes' (/azureml-envs/responsibleai/lib/python3.8/site-packages/sklearn/utils/fixes.py)

It is using the prebuilt RAI environment via a pipeline job. Not sure what other info is needed as this isn't a local issue.

### Steps to reproduce

Submit a pipeline job to create an RAI dashaboard.

### Expected behavior

I would like it to work.

### Actual behavior

it doesn't!

### Addition information

_No response_

Start by reproducing the pipeline job that creates an RAI dashboard in Azure ML Studio with the prebuilt RAI environment. Investigate the reported model-loading failure and the incompatible sklearn.utils.fixes import in the environment. Done means the pipeline loads the MLflow model and creates the dashboard without this import error.

Written by the indexing model from the issue text.