Platform For AI:Create and manage TensorBoard tasks

Account and permission requirements

• Alibaba Cloud account: You can use an Alibaba Cloud account to complete all operations without additional authorization.

• RAM user: You need to add a RAM user as a workspace member of certain roles and assign permissions to the roles. For more information, go to the Roles and Permissions page.

Create a TensorBoard task

On the TensorBoard tab, you can create a TensorBoard task to view the model training process on the TensorBoard page. The following section describes the procedure.

1. Go to the Distributed Training Jobs page

   1. Log on to the PAI console.

   2. In the left-side navigation pane, click Workspaces. On the Workspaces page, click the name of the workspace that you want to manage.

   3. In the left-side navigation pane, choose AI Computing Asset Management > Jobs to go to the Distributed Training Jobs page.

2. On the TensorBoard tab, click Create TensorBoard.

3. In the Create Tensorboard dialog box, configure the required parameters and click OK.

   Parameters:

   • Datasets: the dataset that you specified when you created the job.

   • Summary Relative Path: the relative directory that stores the checkpoints and event files of the training model.

   It takes about 1 minute to create the TensorBoard task. The created task is in the running state.

4. Find the created TensorBoard task and click View TensorBoard in the Actions column.

   You are redirected to the Tensorboard page on which you can monitor the training process.

Manage TensorBoard tasks

• On the Tensorboard tab, view the created Tensorboard tasks.

• Configure a TensorBoard task so that it stops after a period of time that you specify.

  1. On the TensorBoard tab, find the TensorBoard instance that you want to manage and click Auto stop settings in the Operation column.

  2. In the Auto-stop Settings dialog box, set the Duration parameter.