sue-nm5-monitor

Community

Monitor NM5 SUE runs for status, GPU, and storage

Authordongzhuoyao
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Manually tracking the status of scale-up ML/HPC experiments running on the NM5 sandbox requires querying multiple separate systems for Slurm job state, GPU utilization, storage quotas, and experiment progress, leading to fragmented, time-consuming checks. This Skill automates the entire monitoring workflow end-to-end.

Core Features & Use Cases

  • End-to-end NM5 run monitoring: Automatically pulls Slurm scheduler state, GPU utilization, storage/inode usage, and experiment progress for SUE runs on the NM5 sandbox.
  • NM5-specific adaptation: Handles unique NM5 constraints including offline W&B mode, CUDA GPU probing, Slurm partition routing, and strict workspace root validation.
  • Use case: When you submit a scale-up training run to NM5, use this Skill to get a single comprehensive status report with job progress, GPU-hour budget, storage headroom, and potential blockers without manual SSH queries.

Quick Start

Use the sue-nm5-monitor skill to generate a full status report for your active NM5 SUE experiment run.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: sue-nm5-monitor
Download link: https://github.com/dongzhuoyao/deepresearch/archive/main.zip#sue-nm5-monitor

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 537,000+ vetted skills library on demand.