Presentation:

https://youtu.be/ImFpuvxltyQ

mlflow_mlop

Machine Learning Model, MLflow for tracking, Flask for user input, Docker for package containerization, AzureDevops for MLOP

To build, Docker needs:

Dockerfile
app.py
requirements.txt

Inside Docker File:

FROM python:3.7.3-stretch
EXPOSE 5000
# Working Directory
WORKDIR /app

# Copy source code to working directory
COPY .  /app/

# Install packages from requirements.txt
# hadolint ignore=DL3013
RUN pip install --upgrade pip &&\
pip install --trusted-host pypi.python.org -r requirements.txt

ENTRYPOINT [ "python" ] 
CMD [ "app.py" ] 

Type in requirements.txt:

flask == 1.1.2
sklearn
scipy
numpy
pandas
matplotlib
seaborn
schedule
jupyter
mlflow
requests
schedule

from command prompt type:

pip install --user virtualenv
python -m venv myenvi
.\myenvi\Scripts\activate

Install requirements:

pip install -r requirements.txt

From command prompt type:

docker build -t testimage .
docker images --all
docker run --name testcontainer -p 5000:5000 testimage

Go to docker app, containers/Apps, cick "Open in Browser" Standard Port : 80, 443,5000

Flask:

User input A1 and A2 between number 0 or 1 Model predict and show prediction between 0 or 1, 0 for Not Fraud, 1 for Fraud probability.

MlFlow

import mlflow
import mlflow.sklearn
remote_server_uri = "http://0.0.0.0:5000"
mlflow.tracking.get_tracking_uri()
exp_name = "evaluate_metric"
mlflow.set_experiment(exp_name)
mlflow.start_run():