Basics#

In this section, you’ll learn how to use the basic building blocks of Flyte using flytekit. flytekit is a python SDK for developing flyte workflows and task and can be used generally, whenever stateful computation is desirable. flytekit-developed workflows and tasks are completely runnable locally, unless they need some advanced backend functionality like starting a distributed spark cluster.

In this section we’ll take a look at how to write flyte tasks, compose them together to form a workflow, and then read, manipulate and cache data.

Hello World

Hello World

Hello World
Tasks

Tasks

Tasks
Workflows

Workflows

Workflows
Imperative Workflows

Imperative Workflows

Imperative Workflows
Add Docstrings to Workflows

Add Docstrings to Workflows

Add Docstrings to Workflows
Launch Plans

Launch Plans

Launch Plans
Flyte Decks

Flyte Decks

Flyte Decks
Caching

Caching

Caching
Run Bash Scripts Using ShellTask

Run Bash Scripts Using ShellTask

Run Bash Scripts Using ShellTask
Reference Task

Reference Task

Reference Task
Working With Files

Working With Files

Working With Files
Working With Folders

Working With Folders

Working With Folders
Named Outputs

Named Outputs

Named Outputs
Decorating Tasks

Decorating Tasks

Decorating Tasks
Decorating Workflows

Decorating Workflows

Decorating Workflows
Cache Serializing

Cache Serializing

Cache Serializing

Gallery generated by Sphinx-Gallery