Resources step

The first wizard step is Resources. The heading on the page is Training cluster. You can change the cluster on later steps. Prices are per hour across all nodes.

What you see

Pick a cluster in this order:

• Number of nodes: 1 node, 2 nodes, 4 nodes, and so on

• GPU class, for example L4

• GPU size, for example 24 GB

• Number of GPUs, for example 1 GPU

If the organization has BYOC clusters, Compute appears first so you can choose Arena or a registered cluster.

A strip under the pills shows GPUs, CPU, and RAM per node. A new draft starts with the first valid combination.

Moving forward

Next stays disabled until a cluster is chosen. Missing compute shows a blocked banner: Please choose a preferred compute resource.

Save appears only on Summary (with Train or Resume). Earlier steps persist when you click Next or after idle autosave. A cluster is still required before Train on Summary.

On Summary, Train only appears when the experiment is ready to schedule (including a chosen cluster). The tooltip on Train says training usually starts within about ten minutes once resources are ready.

Changing resources later

On steps after Resources, use Edit resources. That opens Select training cluster, with node count and a Resource class dropdown.

After this step

Environment is next. Classic projects go straight to gym selection. Advanced Training projects pick a gym-style RL Environment or a Dataset from the combined grid.

See Environment step.