fix(configs): pin head unschedulable — train (batch 4)#911
Merged
Conversation
Contributor
Author
|
/test-template deepspeed_finetune pytorch-fsdp ray_train_workloads |
Contributor
Author
|
/test-template distributing-pytorch |
Aydin-ab
force-pushed
the
fix/batch-4-head-pin-deps
branch
2 times, most recently
from
July 20, 2026 22:53
6e57695 to
5af157c
Compare
deepspeed_finetune, pytorch-fsdp, ray_train_workloads have CPU-only heads — pin head_node.resources CPU: 0 so work runs on the GPU worker groups. distributing-pytorch's head was a GPU instance only so the Step-1 single-machine baseline could use a local GPU; switch it to a CPU-only coordinator (m5.2xlarge / n2-standard-8, CPU: 0) and run Step 1 on a GPU worker via a Ray task. Claude-Session: https://claude.ai/code/session_01QmLc4yWtmC3PX2NwWzPbzD Signed-off-by: Aydin Abiar <aydin@anyscale.com>
Aydin-ab
force-pushed
the
fix/batch-4-head-pin-deps
branch
from
July 22, 2026 00:29
5af157c to
472dd15
Compare
Contributor
Author
|
/test-template distributing-pytorch |
elliot-barn
approved these changes
Jul 22, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Pins the head unschedulable (
CPU: 0) for the batch-4 train templates.deepspeed_finetune,pytorch-fsdp,ray_train_workloadshave CPU-only heads.distributing-pytorch's head was a GPU instance only so its Step-1 single-machine baseline could use a local GPU — switch it to a CPU-only coordinator and run Step 1 on a GPU worker via a Ray task.Testing
/test-templateon all four (distributing-pytorch's notebook refactor validated on GPU CI, buildkite #533 as #922 + re-run here).Notes
distributing-pytorch was briefly standalone (#922, now closed); folded back to keep batch 4 in one PR.
https://claude.ai/code/session_01QmLc4yWtmC3PX2NwWzPbzD