Starting a notebook server in Watson Machine Learning Accelerator

Edit online

Start a notebook server from the IBM Watson® Machine Learning Accelerator user interface.

Before you begin

Make sure you are familiar with writing notebook server code with TensorFlow or PyTorch.

Procedure

  1. Log in to the Watson Machine Learning Accelerator console as a Cloud Pak for Data project administrator or user.
  2. Click Workload > My Notebook Server.
  3. Click Start server.
  4. Verify server details and click Start. This will start the JupyterLab console.
  5. From the JupyterLab console, select the notebook that you want to start.
    Options include:
    • Python3 with PyTorch Elastic GPU Kernel
    • Python3 with TensorFlow Elastic GPU Kernel
    • Python3 with Single CPU Kernel
    • Python3 with Single GPU Kernel

Results

You should be on JupyterHub code editor so that you can write and deploy your code.