Code does not take latest python version in Azure ML Pipelines

Question

Hi

We have an existing pipeline that was running in production. Recently there were some changes to Python SDK , so we need to change the Python Version from 3.6 to 3.8.

Environment was created with latest version and Published using CI. Below is environment Yaml file

name: py_env
dependencies:
  - python=3.8.13
  - pandas
  - statsmodels
  - scikit-learn
  - pip:
      - azureml-defaults

And we executed the above YAML file using following code

            env_created=Environment.from_conda_specification(env_name, file_path=file_path)

			env_created.register(workspace=ws)

Environment is created in Azure ML Studio, with the above specifications.

We published the pipeline using above Environment Name through CI

But when running the Published Pipeline from Data Factory through endpoint, the Python Version still seems to be 3.6 instead of 3.8.13.

Endpoint is updated, Environment is updated....but still it does not take latest python version.