[Docs] Add the Slurm agent example usage

[JiangJiaWei1103]

We propose to add this example to demonstrate how to use SlurmFunctionTask of the Slurm agent.

Docs link

• Environment setup: https://flyte--6231.org.readthedocs.build/en/6231/deployment/agents/slurm.html
• Example usage: https://flyte--1781.org.readthedocs.build/projects/cookbook/en/1781/auto_examples/slurm_agent/slurm_agent_example_usage.html

Related PRs

• SlurmTask and SlurmShellTask: Slurm agent shell task flytekit#3159
• SlurmFunctionTask: Slurm agent fn task flytekit#3150
• Original: Slurm agent flytekit#3005 (review)

[JiangJiaWei1103]

For agent local test, executing the DL training workflow produces the following result on the Flyte client’s local machine:

[Future-Outlier]

Can you help me?

1. update README.md
2. put training example here
3. use make fmt to fix the lint error

[JiangJiaWei1103]

Can you help me?

1. update README.md
2. put training example here
3. use make fmt to fix the lint error

The training example has been added, and the lint error has been resolved. I’ll complete the README.md tomorrow.

Follow-up

Fix train and run_infer tasks to test GPU usage. I think we've not finished this part as the CUDA device isn't used here, right?

[Future-Outlier]

Can you help me?

1. update README.md
2. put training example here
3. use make fmt to fix the lint error

The training example has been added, and the lint error has been resolved. I’ll complete the README.md tomorrow.

Follow-up

Fix train and run_infer tasks to test GPU usage. I think we've not finished this part as the CUDA device isn't used here, right?

the CUDA is used there, I've done some configuration in the slurm cluster.

[Future-Outlier]

Can you help me?

1. update README.md
2. put training example here
3. use make fmt to fix the lint error

The training example has been added, and the lint error has been resolved. I’ll complete the README.md tomorrow.

Follow-up

Fix train and run_infer tasks to test GPU usage. I think we've not finished this part as the CUDA device isn't used here, right?

not sure why the register CI fails, can you help me deep dive into it?

[JiangJiaWei1103]

No problem, I'll look into this issue tmr!