[SETUP-BUG] How to specify python_version when in log_model i am using pip_requirements?

[dinjazelena]

System information

DBR 13.3 LTS ML

Code to reproduce issue

reqs = {"scikit-learn==1.0.2", "pandas==1.5.2"}
mlflow.pyfunc.log_model(artifact_path="artifacts", python_model=TestWrap(), registered_model_name="wrapperForENV7", await_registration_for=0, pip_requirements=reqs)
model = mlflow.pyfunc.spark_udf(spark, model_uri="models:/wrapperForENV7/latest", env_manager="virtualenv", result_type="double")

# So this one works and gives me this:

2023/11/06 18:42:03 INFO mlflow.pyfunc: This UDF will use virtualenv to recreate the model's software environment for inference. This may take extra time during execution.
2023/11/06 18:42:03 INFO mlflow.models.flavor_backend_registry: Selected backend for flavor 'python_function'
2023/11/06 18:42:03 INFO mlflow.utils.virtualenv: Installing python 3.10.12 if it does not exist
2023/11/06 18:42:03 INFO mlflow.utils.virtualenv: Creating a new environment in /local_disk0/.ephemeral_nfs/repl_tmp_data/ReplId-5a181-004b6-6844f-f/mlflow/envs/virtualenv_envs/mlflow-2b3c092ae9a19d62d8311bd12fd4451d8c54ff5b with /local_disk0/.ephemeral_nfs/repl_tmp_data/ReplId-5a181-004b6-6844f-f/mlflow/envs/pyenv_root/versions/3.10.12/bin/python
2023/11/06 18:42:04 INFO mlflow.utils.virtualenv: Installing dependencies
2023/11/06 18:42:49 INFO mlflow.utils.environment: === Running command '['bash', '-c', 'source /local_disk0/.ephemeral_nfs/repl_tmp_data/ReplId-5a181-004b6-6844f-f/mlflow/envs/virtualenv_envs/mlflow-2b3c092ae9a19d62d8311bd12fd4451d8c54ff5b/bin/activate && python -c ""']'

Describe the problem

How can i specify python_version within requirements.txt so pyenv can install it? Or how is it even possible to specify python_version at all if its not conda.yaml? Since i am using virutalenv as env manager? How to pass there python_version?

Other info / logs

No response

[github-actions]

@mlflow/mlflow-team Please assign a maintainer and start triaging this issue.