Plans
Basic considerations
You can choose one of the following Datalore plans:
Community
Professional
Enterprise
The plans define what computation time, storage options, instances, and extra features are available to you. The information in the sections below will help you choose the right option.
Plan comparison
Refer to this table to compare Datalore plans and explore your options.
Machine usage
With regards to machine usage, it is important to consider the following aspects: computation time, parallel computation, and long computation.
Computation time
Computation time is the time required by the machine to execute the code. To view your current running machines, click your avatar icon in the upper right corner of the screen to open the Account menu and select Running machines.
For a detailed report on machine usage, go to . You can download it as a .csv file from there.
Parallel computation
With the Community plan, you can run 2 notebooks in parallel within the available computation time quota.
With the Professional plan, you can have unlimited parallel computations within the available computation time quota.
Background computation
This mode keeps the machine running after the tab is closed. The option allows you to close your notebook at any point without losing your computation progress. Background computation is particularly helpful when training heavy Machine learning and Deep learning models.
The Community plan allows you to keep your machine running for up to 6 hours after the tab is closed. The Professional plan sets no limitations other than your computation time quota.
Storage
You have the following storage options:
Internal storage
Datalore provides cloud storage for notebooks and attachments. The attached files remain attached to the notebooks whenever you close Datalore. For a .csv report on storage usage, go to .
If you downgrade from the Professional plan to Community, any data exceeding the storage limit will be deleted after 30 days.
External storage
Datalore supports external S3 buckets. This helps you work with the data files you already have in your cloud storage and extend Datalore internal storage based on your needs. To connect your storage, go to .
Sharing
For shared notebooks, Datalore uses the computation resources from the account of the document’s owner. This means that if you are the owner of a notebook that you share with two people and they continue running the notebook after you close the Datalore tab, it is your computation time and memory that will be consumed. If all three of you are running one notebook simultaneously, only one computation will be enabled and consumed.
When you share a workspace, your storage resources will be consumed.
When you publish a notebook, you publish a static copy, and no computation resources will be consumed when the user views it.
Machines
We run your computations on Amazon AWS EC2 virtual servers.
Suitable for | AWS name | Details | |
|---|---|---|---|
Basic machine | Simple data analysis and machine learning tasks | t2a.medium |
|
Large machine | Tasks with huge datasets | r5.large |
|
GPU machine | Deep learning tasks | g4dn.xlarge |
|
For more information, go to Amazon Web Services.