Notebooks are a web-based Jupyter IDE with shared persistent storage for long-term development and inter-notebook collaboration, backed by accelerated compute.
Jupyter kernels are accessible through familiar IDEs like Visual Studio Code and PyCharm. This allows access to the compute connected to the kernel and command line access to the file system from within the IDE. This is not SSH access, so it does not integrate natively with the IDE file manager.
Click the Access Remote Kernel button to access the URI needed for remotely connecting in the IDE.
To run the remote kernel in Visual Studio Code, install the Jupyter extension from the Extensions panel. From the command palette, search for
Specify local or remote Jupyter server for connections.
Then, choose an
Existing kernel and paste the URI from Gradient.
You can confirm you are connected to the remote kernel by confirming you have access to the kernel file system by running
!ls / in Jupyter notebook cell.