PREVIEWYandex DataSphere

Yandex DataSphere is a machine learning (ML) development environment that combines the familiar Jupyter® Notebook interface, serverless computing technology, and seamless use of different computing resource configurations.
Yandex DataSphere helps significantly reduce the cost of machine learning compared to computing on your own hardware or other cloud platforms.
  • Jupyter® Notebook interface
    Work in Yandex DataSphere with a familiar interface. Use Jupyter® Notebook for complex computations, training neural networks, big data analysis, and working with analytical reports.
  • Serverless environment
    Work in a familiar ML development environment, select the necessary computing resource configurations, and run code in specific cells without having to configure servers and virtual machines.
  • Popular libraries
    DataSphere already has pre-installed TensorFlow, PyTorch*, and other major libraries for data analysis and machine learning. Add your libraries for building models, and if these libraries are widely used we’ll add them to the list of the most popular pre-installed libraries.
  • Flexible resources
    Migrate your project to a more powerful computing resource configuration while preserving the environment, variables, code, and data. In a running project, execute different parts of code on the CPU or GPU by changing the configuration in just seconds and without stopping calculations, all while saving the result.
  1. *
    TensorFlow is a registered trademark of Google Inc.
    PyTorch is a registered trademark of Facebook, Inc.