SKlearn_to_MLFLow
In this Repo a simple Sklearn Model will be trained and pushed to MLFlow
Install
This Repo is based on poetry
python3 -m venv .venv
# switch manually to virtual environment and then
$(.venv) poetry install
# will install all dependencies from the pyproject.toml file
General_training
The general Traning pipeline script is configured by the .yaml
file
The yaml file needs to be modified, since the path to the data is hardcoded there.
Data Preprocessing
The data gets prefiltered by the filter.py
configured by filter_config.yaml
in the folder filter and gets stored in the data
folder from where it gets loaded.
ENV File
I have MLFLow in a Docker Container running with a Azurite and Postgres instance. Therefor I need a connection string to connect to MLFlow containers. This is in the .venv
file as well as the connection to MLFlow.
I will link later the Repo for the MLFlow Docker Container.