Concepts

• On-demand Instance: On-demand instances are machines that are available with no long term commitments and offer complete control over the lifecycle of the machine. Use this while running time-critical workloads.
• Temporary Instance (spot / preemptible): Temporary Instances are machines that are available with no long term commitments. They can be interrupted at any point in time. They are usually 30 - 50% cheaper and can be used for not-so critical tasks.
• Project Directory: Directory where you keep all your project-related files such as Jupyter notebooks, python files, temporary checkpoint, and inference output files, etc.
• Environment (docker): The container that is currently running on your machine. Segmind provides base docker containers as templates to start with. Your docker images are automatically saved (and pushed to a registry upon request). You can also bring your own docker as base docker containers. Contact us to know more about bringing your own containers.
• Jupyter (Notebook / Lab): Jupyter Notebook and Lab are two similar IDE for working on data science projects. It offers an interactive interface to code.

Notebook for Enterprise

• Cloud Service Provider (CSP): You can choose which IaaS to use to run your workloads.
• Notebook Region: You can choose which region to run your workloads in. They differ based on CSP.