💥💥My complete journey of MlOps(Machine Learning + Operations) Training 🔥🔥
Aloha, I am Mayank Varshney, student at Uttaranchal Institute Of Technology, pursuing B.Tech in Computer Science having knowledge in Latest and High-End Technologies like Machine Learning, Artificial Intelligence, Deep Learning, MLOps, Dockers, Hybrid-Cloud and many more on the way…
◆ MY BELIEF
“No technology is challenging or difficult or complex as the world says to us or approaches to us since it is MAN-MADE. You just need the right path or approach to understand the technology and take it further as per one’s requirement”. Right thinking or positive thinking that “Yes I Can Do It” can change one’s life drastically.
~MR. VIMAL DAGA (my mentor/teacher)
To be very honest, I have never written a self-reflection blog before, this is my very first time i am writing my own thoughts. This is only because of this 45 days of my MlOps journey had changed my personality, my thinking power, and many more great changes, so please do read this short blog once as you'll get to know the right meaning of #righteducation at the end.
◆ What I learn during the whole journey :-
Step 1:- Learn about RHEL( Red-hat Enterprise Linux)
At the very starting, we get to know about the Redhat OS and its integration with python3, as it is related to our mlops training so we learn about Linux Os from basic to advanced.
Step2:- Start of MlOps Learning
#Computer Vision
Basically we learn the right meaning of CV in this training, we get to know how to work properly on images, here we learn various packages from zero to advanced like numpy, opencv, etc. We got the proper idea of image processing and how to work on it. Learnt the idea of using the Haarcascade for face detection also worked on pretrained models for testing and learning the real-world use cases, like yolo etc.
#Machine learning
Before the starting of the actual Ml concept we got to know the basic idea behind about why to use Ml , what is its need in the industry and what are its future scope. Further we start learning by knowing about the basic libraries like numpy ,pandas, matplotlib, learn how to read the dataset, use of pandas library, use of numpy array, work on matrices and learn matplotlib to visual the graph, learn Linear regression, learn training and testing concept of dataset, learn how to fit/train the model and how to predict the model.
Most importantly what we learn here is that we can train our model once and then we can predict multiple outputs with different use cases.
learn linear regression algorithm, got to know the main idea of linear regression concept, learn how to find the absolute error and min error concept, scikit-learn algorithm, optimization concept of algorithm and multi linear regression algorithm.
Learn tensor-flow and keras algorithm, lazy execution, distributing computing, Deep learning concept perception, neuron, Gradient descent, auto feature selection, activation function relu, sigmoid, Threshold, tanh etc..Discuss feed forward ,forward propagation, Back Prorogation, learning rate concept, epoc concept, gradient descent, optimizer Adam etc, data visualization, Matplotlib, seaborn, pandas, etc.
Most importantly we learn Folium library to create Maps where we learn adaptive learning rate.
step3:- Move forward to Deep Learning world
#Deep Learning
In this we learn about basics of perception, input layer, hidden layer, output layer, discussed single neuron and multi neuron, about optimizer, dense layer, learn binary cross entropy, Multi classification algorithm, about CNN(convalutional Neural Network) use to eliminate the feature, manually and automatically, VGG, Res-net, inception method of CNN algorithm , Fine tuning and transfer learning.
Transfer learning is most important where we created/used few images to lots of dataset using augment concept.
Learn VGG16, VGG19, Mobile-net, freeze layer concept, KNN Algorithm, clustering algorithm, Mask-R_CNN Algorithm,GAN’S Algorithm, Association rule, apriori algorithm and recommendation system concept, RBM conceptand many more.
Step4:- Started implementing MlOps concept on DevOps world
#Docker
Learn concept of Docker, installation of docker, downloaded the docker images, launch the web-server, use of docker image, docker networking.
Most importantly we learn to create a docker image & use of Docker-file concept.
# Git and GitHub
learn concept of git, git source code controller, learn how to create git repository and push on the GitHub and store , git commit concept , git merge concept, git pull concept etc..
#Jenkins
Learn concept of Jenkins, copy the code on the GitHub.SCM concept, Poll SCM concept ,crone tab concept, GitHub, tunnel concept, plugin concept, job chaining, automation etc.
#Kubernetes
Learn concept of Kubernetes, installation of Kubernetes, how to launch pods, about minikube, replica set, yaml programming language and cli command of kubernetes, learn concept of Replication controller, concept of replica set, etc…
Here we used the actual Machine Learning with Operations, where we created multiple tasks that will automate the Ml environment .
Solve some task on this MlOps journey:-
At last, I would like to thanks Mr. Vimal Daga Sir and LinuxWorld India Pvt. Ltd. for providing us with this wonderful opportunity, i have never thought that i can do such great things in the computer world, all this happens because of your #righteducation and your vision of #makingindiafutureready , also i would like to thanks Preeti ma'am who helped us a lot during the whole journey and also taught us the marketing skills by giving us the ambassadors opportunity.
