Skills
.Work. You
Rest

flyte-kuberay

Community

Run Ray workloads as Flyte tasks via KubeRay

Data & Analytics#kubernetes#ray#distributed-training#workflow-management#flyte#kuberay#ray-job
Authortylertitsworth
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Seamlessly run Ray workloads as Flyte tasks by provisioning ephemeral Ray clusters per task execution via the KubeRay operator. The integration enables Flyte users to configure RayJobConfig (head and worker resources, autoscaling, runtime_env) and run Ray Core, Ray Data, Ray Train, or Ray Serve within Flyte workflows while ensuring clusters are created, used, and torn down per task run.

Core Features & Use Cases

  • Ephemeral Ray clusters per task: Flyte launches a dedicated Ray cluster for each execution and tears it down after completion.
  • Rich Ray integration: Supports Ray Core, Ray Data, Ray Train, and Ray Serve within Flyte tasks.
  • Flexible configuration: Customize head and worker node settings, autoscaling, pod templates, and runtime environments via RayJobConfig.
  • Safe operational flow: Manages per-task lifecycle and resources in a Kubernetes environment.

Quick Start

Run a Flyte task that uses the flytekitplugins-ray integration to submit your Ray function to an ephemeral cluster.

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: flyte-kuberay
Download link: https://github.com/tylertitsworth/skills/archive/main.zip#flyte-kuberay

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 223,000+ vetted skills library on demand.