Yandex Cloud
Search
Contact UsGet started
  • Blog
  • Pricing
  • Documentation
  • All Services
  • System Status
    • Featured
    • Infrastructure & Network
    • Data Platform
    • Containers
    • Developer tools
    • Serverless
    • Security
    • Monitoring & Resources
    • ML & AI
    • Business tools
  • All Solutions
    • By industry
    • By use case
    • Economics and Pricing
    • Security
    • Technical Support
    • Customer Stories
    • Gateway to Russia
    • Cloud for Startups
    • Education and Science
  • Blog
  • Pricing
  • Documentation
Yandex project
© 2025 Yandex.Cloud LLC
All solutions
    • All solutions for DataSphere
    • Fixing project startup errors
    • Fixing the "Failed to deserialize variable" error
    • You do not see the Table of Contents extension in the JupyterLab extension list
    • Getting access to the GPU g1.1 and higher configurations in DataSphere
    • Resolving issues with long code execution within DataSphere project cells
    • Resolving the "Servant not allocated" error when running code within DataSphere project cells
    • Resolving the "TestsFailure" error
    • Resolving the project opening error
    • Resolving the "Token authentication is enabled" error
    • Resolving errors with project access or login
    • Resolving the "KernelNotResponding The kernel died unexpectedly" error

In this article:

  • Issue description
  • Solution
  • If the issue persists
  1. DataSphere
  2. Fixing the "Failed to deserialize variable" error

Fixing the «Failed to deserialize variable» error

Written by
Yandex Cloud
Updated at September 12, 2023
  • Issue description
  • Solution
  • If the issue persists

Issue descriptionIssue description

When you open a DataSphere project, you see the following error message:

Failed to deserialize variable %VAR_NAME%
Run the following code to delete it: del_datasphere_variables

Where %VAR_NAME% is the name of the variable that you couldn't deserialize.

SolutionSolution

The message The following variables cannot be serialized means that some variables fail to be serialized, so you cannot save them.
The interpreter state will be saved on configuration change. All the computation results, including values of variables, will also be saved at this point.

Some variables fail to be serialized and cannot be saved. For example, a variable with a file open for writing: f = open("file.txt", "w").
When you assign values to such variables, you'll see the warning: The following variables cannot be serialized.
Such a warning will not affect your code execution and further computation results.

The Run the following code to delete it: del_datasphere_variables message means that the variable was probably stored in the RAM of your GPU configuration previously, but hasn't been moved to the RAM of the current GPU configuration.

The error text includes a recommendation how to reset the variable (the del_datasphere_variables method) to fix the error. In such a case, try running all the cells linked to this variable within the same configuration.

If the issue persistsIf the issue persists

If the above actions didn't help, create a request for support.
In the request, specify the following information:

  1. Project URL from the browser address bar.
  2. Full text of the error message.
  3. HAR file with the recorded results of browser and server interaction with the DataSphere project.

Was the article helpful?

Previous
Fixing project startup errors
Next
You do not see the Table of Contents extension in the JupyterLab extension list
Yandex project
© 2025 Yandex.Cloud LLC