Integration of ML with Docker

So, here are some steps for the deployment of ML model on Docker:-

Step 1:- Pulling CentOS image from DockerHub

docker pull <image_name>

Step 2:- Launching new Container

docker run -it --name <container_name> <image_name>

Step 3:- Installing Python in newly launched Container

yum install python3

Step 4:- Copying ML code in Container from BaseOS

Here, the model is trained and stored in score.pk1 file and in score.py file, it is loaded to find the prediction value. So, both the files are being copied in the container.

docker cp <src_file> <container_name>:<dest_path>

Step 5:- Installing required Python Libraries

Before we run the model, we have to install the scikit-learn(provides sklearn module) and joblib(provides joblib module) library. sklearn module is used for training the model and the joblib module is to load the model(score.pk1) in our code.

pip3 install <lib_name>

Step 6:- Running program

python3 <py_file>

Thank You

Hope you find it helpful...