[None][docs] drop ghost enable_overlap_scheduler from trtllm-serve Slurm example - #18961
Conversation
…urm example The Multi-node Slurm sample nested pytorch_backend_config.enable_overlap_scheduler, which is not a TorchLlmArgs field. Overlap is on by default; live knob is top-level disable_overlap_scheduler (see docs/source/features/overlap-scheduler.md). Keep enable_attention_dp: true; drop the ghost overlap stanza. Signed-off-by: ADou <ikun3.1415927@gmail.com>
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
Included review availability: Your plan provides up to 12 included reviews per hour; 7 remain after this review. WalkthroughThe Slurm multi-node serving example now writes only ChangesSlurm serving configuration
Priority: ⬇️ Low Estimated code review effort: 1 (Trivial) | ~2 minutes Merge Risk: ⚪ Minimal · up to The Slurm serving example no longer documents an unsupported overlap-scheduler option and continues to enable attention data parallelism. No current merge-readiness risk remains. Suggested reviewers: 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Summary
docs/source/commands/trtllm-serve/trtllm-serve.rstwrotepytorch_backend_config.enable_overlap_scheduler: true.TorchLlmArgsfield; the live knob is top-leveldisable_overlap_scheduler(defaultfalse, i.e. overlap on). Seedocs/source/features/overlap-scheduler.md.enable_attention_dp: truein the sample YAML; drop the ghost overlap stanza.Test plan
enable_overlap_schedulerhas 0 Python hits on main; only this rst referenced itdisable_overlap_scheduleris the documented/live field inllm_args.pyand feature docsDev Engineer Review
The Multi-node Slurm example removes the unsupported
pytorch_backend_config.enable_overlap_schedulersetting. It retainsenable_attention_dp: trueand relies on the supported default that enables overlap scheduling.QA Engineer Review
No test changes.
Per-File QA Perspective
docs/source/commands/trtllm-serve/trtllm-serve.rst: Verify that the example generates valid configuration and preserves attention data parallelism and default overlap scheduling behavior.