MLOps vs. DevOps: What is the Difference?
(A thread) 👇
(A thread) 👇
MLOps emphasizes the need for collaboration between data scientists and engineers early on in the process, whereas DevOps focuses on automating the software development process.
The key difference is the requirement of data versioning along with the source code in ML systems.
The key difference is the requirement of data versioning along with the source code in ML systems.
MLOps can be thought of as a branch of DevOps that focuses on the collaboration and integration of ML models into software development process.
DevOps = Code
MLOps = (Data + Code)
Data and code is like Soul and Body, both of them evolve independently but are still connected.
DevOps = Code
MLOps = (Data + Code)
Data and code is like Soul and Body, both of them evolve independently but are still connected.
DevOps focuses on automating the software development and delivery process.
In most of the standard DevOps setup, Jenkins is used in collaboration with git to efficiently build, test, and deploy versioned code in a controlled environment.
In most of the standard DevOps setup, Jenkins is used in collaboration with git to efficiently build, test, and deploy versioned code in a controlled environment.
MLOps is about innovatively applying the existing DevOps practices to automate the building, testing, and deployment of large-scale machine learning systems. In the case of MLOps, the CI/CD workflows get triggered either by a change in source code or data.
The goal of MLOps is to perform continuous training of the model by automating the entire ML pipeline that leads to continuous delivery of prediction service.
What powers continuous model training is the ability to do data version control along with code.
What powers continuous model training is the ability to do data version control along with code.
Here are some innovative companies leading the MLOps revolution from the front:
- Weights and Biases
- Jina AI
- Comet ML
- Neptune AI
- Fiddler AI
- Dataiku
- Zen ML
- Weights and Biases
- Jina AI
- Comet ML
- Neptune AI
- Fiddler AI
- Dataiku
- Zen ML
That's a wrap!
If you liked this thread and want to access more such ML and Data Science content, consider following me @Saboo_Shubham_🙏
If you liked this thread and want to access more such ML and Data Science content, consider following me @Saboo_Shubham_🙏
Loading suggestions...