lightning-jobs

Orchestrate batch jobs and distributed training on the Lightning AI platform.

5|Updated Jul 3, 2026

Search Tags:#automation #gpu #distributed-training #batch-jobs #cloud-computing #lightning-ai

Ready to run

npx skills add https://github.com/Lightning-AI/skills --skill lightning-jobs

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: lightning-jobs
Source: https://github.com/Lightning-AI/skills/tree/main/lightning-jobs
Command: npx skills add https://github.com/Lightning-AI/skills --skill lightning-jobs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires lightning-sdk, jq.

What problem does it solve?

This skill removes the complexity of managing cloud infrastructure for batch workloads, allowing you to run training, data processing, or distributed tasks without manually configuring servers.

Core Features & Use Cases

Flexible Execution: Run jobs inside custom Docker images or snapshots of existing Lightning Studios.
Distributed Training: Easily launch multi-machine (MMT) jobs for large-scale distributed training.
Monitoring & Management: Stream logs, inspect job status, and manage artifacts directly through the CLI or SDK.

Quick Start

Use the lightning-jobs skill to launch a new batch job on an L4 GPU using the pytorch image and run the training script.

Frequently Asked Questions about lightning-jobs

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

FAQPage Schema

How do I launch distributed training jobs across multiple GPU machines?▼

To launch distributed training jobs, you can use the lightning-jobs skill to orchestrate multi-machine tasks across CPU and GPU clusters, managing the lifecycle of large-scale batch workloads without manual server configuration. It supports programmatic interfaces for execution and monitoring.

What is the best way to run batch jobs using a custom Docker image in the cloud?▼

Running batch jobs with custom Docker images is supported directly through the lightning-jobs skill, which orchestrates containerized image execution on the Lightning AI platform. This removes the complexity of manually provisioning and configuring cloud infrastructure for your workloads.

How do I monitor cloud batch job status and stream execution logs?▼

You can monitor cloud batch jobs and stream execution logs directly through the lightning-jobs skill using either the command line interface or the software development kit. It provides programmatic interfaces for job monitoring, log streaming, and artifact collection during task execution.

Can I use a Lightning Studio environment snapshot to run a new batch job?▼

Yes, you can use existing Lightning Studios environment snapshots to run new batch jobs. The lightning-jobs skill supports executing tasks directly from these snapshots, allowing you to seamlessly replicate your configured workspace environments for cloud-based batch processing and distributed training.

Do I need lightning-sdk and jq installed to manage batch jobs?▼

Yes, you need both lightning-sdk and jq installed in your local environment to use the lightning-jobs skill. These dependencies are required to interface with the Lightning AI platform and process the JSON data structures returned during batch job orchestration and monitoring.