Скачать или смотреть Complete End-to-End building and deploying ML Regression model with FastAPI, MLFlow, Docker, & AWS!🔥

In this video, I share how I built and deployed an end to end ML project using production grade code, best practices, tools and documentation.

The goal was not to build the best model.
But to emulate the complete machine learning life cycle.

1. Model building (EDA, cleaning, feature engineering, baseline model creation, trying out different models - Linear, lasso and ridge etc. → XGBoost)
2. Finetuning the Model using Optuna and logging all experiments using MLflow.
3. Creating HTTP endpoint using FastAPI for backend.
4. Creating frontend using Streamlit.
5. Containerizing both(frontend & backend) using Docker.
6. Implementing CI/CD pipelines using GitHub Actions.
7. Deploying everything using AWS—used services like:
‣ S3 - for storing data & model
‣ ECR - to store docker images
‣ ECS - to run the containers (task definition + setting security and target
group + execution role)
‣ Application Load Balancers - to get public url!
8. Finally pushing everything up!!

Link to Notion notes - https://www.notion.so/Build-Deploy-ML...
Link to GitHub - https://github.com/aditya-somani/End-...

Connect with me on Linkedin -   / aditya-somani-03631727b