Duration
MLOps Training
3 days
Overview
As complexity of Machine learning projects increases there is a need to manage the end to end lifecycle of such AI ML projects. This termed as MLOps is mandatory for all ML projects. We offer hands on workshop on covering all aspects from development to operations of ML with focus on industry leading MLOps tool MLFlow.
Course Outline
- ML life cycle with crisp-dm
- Introduction to MLOps
- Role of MLOps In ML Lifecycle
- MLOps tools landscape
- MLflow as MLOps
- Components of MLOps tools
- Experiments
- Tracking
- Model management
- Registry
- Versioning
- Hyper parameter tuning
- Data processing, Wrangling and Feature Engineering
- Basic Cleaning Steps for Data
- Dimensionality Reduction
- Feature Selection
- Feature Engineering
- Model development & management with MLFlow
- Common Steps
- Model Development
- Hyperparameter Tuning
- Model Evaluation
- Supervised ML
- Supervised Classification
- Supervised Regression
- Unsupervised ML
- Unsupervised clustering
- Continuous Integration and Deployment with MLFlow
- Model Deployment with Flask
- Using Experiments and Runs
- Model Registry
- Model Monitoring and Model Evaluation
- Model Evaluation
- Model Monitoring
- Best practices End to End Model Lifecycle
Programme scope, delivery arrangements and suitability are confirmed during enquiry.