-
Book Overview & Buying
-
Table Of Contents
-
Feedback & Rating

Computer Vision on AWS
By :

Now that we’ve discussed the best practices to apply to each stage of the ML life cycle, how can we automate and streamline these processes? We can accomplish this by incorporating MLOps. What is MLOps? MLOps is related to DevOps in concept, where both practices focus on automating and accelerating applications or systems from development to production. The difference between the two is that the goal of DevOps is to deliver software applications, while the goal of MLOps is to deliver ML models. MLOps allows you to automate your ML workflows and create repeatable mechanisms to accelerate the processes for building, training, deploying, and managing ML models. You can leverage tools such as workflow automation software for orchestration and continuous integration/continuous delivery (CI/CD) of your ML systems. Other components of MLOps include tracking lineage using a Model Registry. Also, monitoring models in production and providing corrective actions...