All flytekitplugins maintained by the core team are added here. It is not necessary to add plugins here, but this is a good starting place.
| Plugin | Installation | Description | Version |
|---|---|---|---|
| AWS Sagemaker Training | bash pip install flytekitplugins-awssagemaker |
Installs SDK to author Sagemaker built-in and custom training jobs in python |  |
| Hive Queries | bash pip install flytekitplugins-hive |
Installs SDK to author Hive Queries that can be executed on a configured hive backend using Flyte backend plugin | |
| K8s distributed PyTorch Jobs | bash pip install flytekitplugins-kfpytorch |
Installs SDK to author Distributed pyTorch Jobs in python using Kubeflow PyTorch Operator | |
| K8s native tensorflow Jobs | bash pip install flytekitplugins-kftensorflow |
Installs SDK to author Distributed tensorflow Jobs in python using Kubeflow Tensorflow Operator | |
| Papermill based Tasks | bash pip install flytekitplugins-papermill |
Execute entire notebooks as Flyte Tasks and pass inputs and outputs between them and python tasks | |
| Pod Tasks | bash pip install flytekitplugins-pod |
Installs SDK to author Pods in python. These pods can have multiple containers, use volumes and have non exiting side-cars | |
| spark | bash pip install flytekitplugins-spark |
Installs SDK to author Spark jobs that can be executed natively on Kubernetes with a supported backend Flyte plugin | |
Please file an issue and or create a PR.
Flyte plugins are structured as micro-libs and can be authored in an independent repository. The plugins maintained by the core team are maintained in this repository
All plugins should expose a library in the format flytekitplugins-{}, where {} is a unique identifier for the plugin.
Plugins should have their own unit tests