INTEGRATING ML(machine learning) and DevOps

In this, we will discuss how we can integrate ML(machine learning) and DevOps by building a small project.

we will learn about the steps which are required for this integration by building a project on for writing code on image recognition using the concept of and then we will integrate it with DevOps using Jenkins for automation of our work.

1.first we will consider a predefined weight of imagenet and use a dataset of coins of different for our model for transfer our learning.

2. after that we will check whether our model is accurate or not and note our output for further work.

3.now here we will Jenkins to create a different job to automate our work of testing.

JOB 1: In this file, older files will be removed and new files will be downloaded by Jenkins.

after the above step, we will first create Docker images for Jenkins so that it can check the images and if it is suitable then use to test the code.

the images are:-

image 1-here we are creating image having Keras, TensorFlow with python3 and we are using centos as base os.

image 2-here we are using another image which has sklearn and pandas as docker hub.

after the creation of images, we will create our Job 2.

Job 2: this job will check the docker images whether they are suitable for our model or not using data file handling [DFH] and then by importing the module in python we will run the code directly from the python file.

Job 3: then we will run our main files for training our model.

Job 4: In this job, we will check the accuracy of our model by running different code for it.

Job 5: This job will only upload the updated code to our repository on Github.

After following al these, we are able to integrate Machine learning with DevOps. this integration is essential for automation of system and this concept is the future of upcoming automation systems.