MLOps (Machine Learning Operation) refers to a set of best practices, tools and technologies that are used to streamline and manage the lifecycle of machine learning models. It combines machine learning (ML) with DevOps (Development and Operations) principles to create a consistent and automated workflow for developing, deploying, and maintaining ML models.
MLOps encompasses various stages of the ML model development lifecycle, including data preparation, model training, model deployment, monitoring, and retraining. It aims to address the challenges associated with ML model development, such as reproducibility, scalability, versioning, and collaboration.
| Topic | Subtopic |
|---|---|
| Desinging ML Systems | MLOps Overview |
| Training System Vs. Inference System | |
| Prediction Patterns | |
| Data Preparation | Active Learning |
| Weak Supervision | |
| Model Development | Model Baseline |