Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[RLlib] Enhancements for multi-node/multi-GPU training and better EnvRunner error msg. #47705

Merged

Conversation

sven1977
Copy link
Contributor

@sven1977 sven1977 commented Sep 17, 2024

Enhancement for multi-node/multi-GPU training and better EnvRunner error msg.

  • --num-gpus>0 now also works on multi-node multi-GPU setups where the head node does NOT have a GPU.
  • add better error messages to EnvRunners in case config.env is None.

Why are these changes needed?

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 :(

Signed-off-by: sven1977 <[email protected]>
@sven1977 sven1977 added rllib RLlib related issues rllib-env rllib env related issues rllib-gpu-multi-gpu RLlib issues that's related to running on one or multiple GPUs rllib-newstack labels Sep 17, 2024
Copy link
Collaborator

@simonsays1980 simonsays1980 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Small nit with an error output.

@@ -1392,15 +1392,18 @@ def run_rllib_example_script_experiment(
# Define compute resources used automatically (only using the --num-gpus arg).
# New stack.
if config.enable_rl_module_and_learner:
# Do we have GPUs available in the cluster?
num_gpus = ray.cluster_resources().get("GPU", 0)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice!

if args.num_gpus > 0 and num_gpus < args.num_gpus:
raise ValueError(
f"You are running your script with --num-gpus={args.num_gpus}, "
"but your cluster only has {resources['GPU']} GPUs! Will "
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Something missing? "Will ... "?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fixed. Now, it'll just warn and then run with CPU-based Learners instead.

Signed-off-by: sven1977 <[email protected]>
@sven1977 sven1977 enabled auto-merge (squash) September 17, 2024 12:23
@github-actions github-actions bot added the go add ONLY when ready to merge, run all tests label Sep 17, 2024
@sven1977 sven1977 merged commit c1e7c45 into ray-project:master Sep 17, 2024
7 checks passed
@sven1977 sven1977 deleted the fix_multi_node_multi_gpu_training branch September 18, 2024 07:57
ujjawal-khare pushed a commit to ujjawal-khare-27/ray that referenced this pull request Oct 15, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
go add ONLY when ready to merge, run all tests rllib RLlib related issues rllib-env rllib env related issues rllib-gpu-multi-gpu RLlib issues that's related to running on one or multiple GPUs rllib-newstack
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants