AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |
Back to Blog
We can drop Mac immediately, because it does not have an option to include NVIDIA GPU, and you need to have it for any serious model training. In terms of operating systems for your local environment, you have a choice of Linux, Windows, and Mac. Having your own machine makes a lot of sense. for compliance or regulatory reasons.Īdditionally, I like to be able to take my work with me on a trip, and sometimes I’m not within range of a fast internet connection. If you are an experienced professional you will always meet a customer which cannot put their data on a public cloud, ie. If you are starting your career it’s better to understand exactly how all the pieces are working together, experiment with many tools and frameworks, and do it in a cost-effective way. Nevertheless, every data scientist needs a local environment. They are also great in terms of productizing your solution and exposing an inference endpoint. You can set up a VM, container, or use a ready-made environment that presents you with a Jupyter notebook. Public clouds offer a great set of solutions for data professionals. I would however recommend reading the reasoning behind certain choices to understand why this is the recommended setup. TLDR If you just want a tutorial to set up your data science environment on Ubuntu using NVIDIA RAPIDS and NGC Containers just scroll down.
0 Comments
Read More
Leave a Reply. |