The simple and easy-to-understand management interface greatly reduces the burden of deployment, operation, and real-time management.
Training and deploying model without writing extra code. The system currently supports common models such as TensorFlow, PyTorch, and SKLearn.
Data Scientist Team can write Dockefile and deploy through PrimeHub Deploy.
Reducing unnecessary downtime while deploying model by rolling update.
Allocating computing resources through real-time data monitoring.
Getting comprehensive Model API deployment status of each project.
Setting different authentication for different Model API deployments
Adjusting Model API deployment computing resources dynamically.
Checking every Model API deployment changelog.
PrimeHub is an on-premise machine learning platform that enables AI/ML teams to focus on their true productivity in a collaborative environment. PrimeHub helps administrators manage hardware resources, access control, group quota, datasets and more.
Administrators/IT Leaders and Data Scientists. Administrators set up environments for the data scientists to use. They can create custom images and allocate resources to each user, then set up each users’ data access and group permissions. Data scientists are given a seamless ML experience with their own customized Jupyter notebook environment.
PrimeHub CE is our single node version of PrimeHub with basic features and is available for anyone to use and contribute to. Visit our GitHub and ensure that you have the prerequisites you need to get started. You can also visit our documentation site for more info.
Yes. With our Job Submission feature, users are able to submit time-consuming jobs to run in the background, set a scheduled recurrence when creating a new job, or choose to run a job manually at a later time.
Administrators have the ability to add users to groups, and to allow groups access to specific data sets using the Admin dashboard.