Which job schedulers are supported by this skill?

This skill is specifically optimized for SLURM (Simple Linux Utility for Resource Management) environments commonly used in academic and research clusters.

Does it work with Weights & Biases?

The skill includes logic to detect offline W&B runs that are pending synchronization, helping you keep your cloud-based experiment tracking up to date.

Can I filter the status report for a specific project?

Yes, you can pass a dataset name as an optional argument to filter the results and receive a detailed status table for that specific experiment set.

How does this skill help with debugging failed experiments?

It automatically identifies and lists the most recent error logs from your SLURM directory and checks for missing checkpoints or configuration files.

ML Experiment & SLURM Monitor

Name: ML Experiment & SLURM Monitor
Author: RobertFrenken

byRobertFrenken

0•

データサイエンスとML

Monitors machine learning experiment progress, SLURM job queues, and model checkpoint status across distributed computing environments.

The check-status skill is a specialized utility for researchers and data scientists working with high-performance computing (HPC) clusters. It automates the tedious task of monitoring multi-stage machine learning pipelines by combining SLURM job queue data, PyTorch checkpoint verification, and error log analysis into a single consolidated view. It is particularly useful for tracking complex DQN-Fusion experiments, providing instant visibility into which datasets have completed training stages and identifying synchronization issues with tools like Weights & Biases.

主な機能

01Weights & Biases (W&B) offline run and sync monitoring

020 GitHub stars

03Automated PyTorch model checkpoint and metrics verification

04Detailed pipeline status reporting for specific datasets

05Recursive log scanning for recent SLURM execution errors

06Real-time SLURM job queue tracking with custom formatting

ユースケース

01Monitoring parallel reinforcement learning training runs on an HPC cluster

02Verifying the integrity of model checkpoints and config files across datasets

03Rapidly debugging failed jobs by accessing recent error logs and queue status

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add robertfrenken/dqn-fusion check-status

For use in Claude.ai and ChatGPT

Download Skill