Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upGitHub is where the world builds software
Millions of developers and companies build, ship, and maintain their software on GitHub — the largest and most advanced development platform in the world.
TFOperator support in FlytePlugins #115
Open
Assignees
Comments
|
cc @swiftdiaries is interested in doing this |
|
Ack :) I'll start work on this next week |
|
@igorvalko recently contributed pytorch-operator plugin for Flyte. The issue is #338 |
|
I have added all of his PRs to the issue - #338 (comment) |
|
https://github.com/pingsutw has volunteered to help with this |
|
Work in progress at - https://github.com/flyteorg/kfplugins |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
TL;DR
Add support for new "tensorflow" task type, that uses TFOperator under the hood.
Summary:
Flyte allows orchestrating across different orchestrators. Though a simple TF job could be python code, using TFOperator allows other benefits. Namely Tensorboard and ability to run distributed tf jobs.
Operator to use:
https://github.com/kubeflow/tf-operator
Components: