[train] reference custom resources in ScalingConfig.resources_per_worker

[matthewdeng]

Why are these changes needed?

Make it clear that you can specify Ray Core custom resources here.

Related issue number

Checks

• I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
• I've run scripts/format.sh to lint the changes in this PR.
• I've included any doc changes needed for https://docs.ray.io/en/master/.
  • I've added any new APIs to the API Reference. For example, if I added a
    method in Tune, I've added it in doc/source/tune/api/ under the
    corresponding .rst file.
• I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
• Testing Strategy
  • Unit tests
  • Release tests
  • This PR is not tested :(

[scottsun94]

Does it support memory without custom resources or only cpu and gpu?

[justinvyu]

Add custom resources to the example below:

from ray.train import ScalingConfig

scaling_config = ScalingConfig(
    # Number of distributed workers.
    num_workers=2,
    # Turn on/off GPU.
    use_gpu=True,
    # Assign extra CPU/GPU/custom resources per worker.
    resources_per_worker={"GPU": 1, "CPU": 1, "memory": 1e9, "custom": 1.0},
    # Try to schedule workers on different nodes.
    placement_strategy="SPREAD",
)

Also:

1. Let's remove this trainer_resources section in the docs, or just keep the section talking about setting it to 0. https://docs.ray.io/en/latest/train/user-guides/using-gpus.html#trainer-resources
2. Link the scaling gpu resource configuration guide at the top of the ScalingConfig docstring.
3. Update the resources_per_worker section to not emphasize setting multiple GPUs. We already have another section about this, and it's not the recommended pattern.

[hongpeng-guo]

LGTM