monitor-experiment

Name: monitor-experiment
Availability: InStock
Author: Chenghao

Monitor ML experiment progress and collect results across remote SSH servers.

2|Updated Mar 19, 2026

Search Tags:#logs #ssh #experiment #monitor #results #wandb #screen

Ready to run

npx skills add https://github.com/chenghaoYang/auto-coder-trainer --skill monitor-experiment-chenghaoyang

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: monitor-experiment
Source: https://github.com/chenghaoYang/auto-coder-trainer/tree/main/aris/skills/monitor-experiment
Command: npx skills add https://github.com/chenghaoYang/auto-coder-trainer --skill monitor-experiment-chenghaoyang

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Monitor running experiments, check progress, and collect results across servers and screen sessions to keep teams informed about ongoing work.

Core Features & Use Cases

Real-time progress checks: verify that experiments are still running and progressing.
Result collection: fetch latest outputs, logs, and JSON metrics from remote hosts.
Status reporting: summarize run state and flag stalled or failed tasks for rapid intervention.

Quick Start

Connect to the monitoring servers via SSH and review the latest experiment progress and results.

Frequently Asked Questions about monitor-experiment

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I track ML experiment progress on a remote server via SSH?▼

To track ML experiment progress on a remote server via SSH, you can connect to the host and review screen sessions to verify that training is still running and progressing in real time.

How do I collect JSON metrics and logs from running experiments across remote servers?▼

Collect JSON metrics and logs from running experiments by connecting to remote servers via SSH, accessing results directories, and fetching the latest outputs to summarize the current run state.

Can I monitor long-running experiments in screen sessions to detect stalled or failed tasks?▼

Yes, you can monitor long-running experiments in screen sessions to detect stalled or failed tasks by checking run states and summarizing status for rapid intervention across servers.

Do I need SSH access to check experiment status and collect results from remote hosts?▼

Yes, you need SSH access to check experiment status and collect results from remote hosts, along with screen session management and access to the directories storing the outputs.

What is the best way to summarize run state and flag failed ML training tasks?▼

The best way to summarize run state and flag failed ML training tasks is to fetch latest logs and JSON metrics from remote servers, summarizing the experiment status to keep teams informed.