Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Tune] Working with checkpoints user guide #32659

Closed
justinvyu opened this issue Feb 17, 2023 · 0 comments · Fixed by #33145
Closed

[Tune] Working with checkpoints user guide #32659

justinvyu opened this issue Feb 17, 2023 · 0 comments · Fixed by #33145
Assignees
Labels
docs An issue or change related to documentation P1 Issue that should be fixed within a few weeks Ray-2.4 ray-team-created Ray Team created tune Tune-related issues

Comments

@justinvyu
Copy link
Contributor

justinvyu commented Feb 17, 2023

This "working with checkpoints" guide has been converted to a Tune storage guide, and the content in the old guide was always more to do with syncing than teaching users how to create/load checkpoints.

The only information about creating trial checkpoints is here in the Tune API ref: https://docs.ray.io/en/master/tune/api/trainable.html

Once this guide is created, remove the URL redirect from tune-checkpoints -> tune-storage.

We should create a user guide called tune-checkpoints again for this common workflow. Here's what it could look like:

How to Save and Load Checkpoints in Ray Tune

Saving

For class + function trainables.

Use this content:

Loading

For class + function trainables.

Configurations

Explaining CheckpointConfig

  • num_to_keep
  • checkpoint frequency for class Trainable vs. an example of how to set it manually for function trainable

See Also

Checkpointing enables:

  • Fault tolerance guide
  • Storage options guide
  • Batch prediction guide

Appendix: Experiment Checkpoints vs. Trial Checkpoints

Put this content in an appendix: https://docs.ray.io/en/releases-2.2.0/tune/tutorials/tune-checkpoints.html#two-different-types-of-tune-checkpoints

@justinvyu justinvyu added tune Tune-related issues P2 Important issue, but not time-critical docs An issue or change related to documentation air Ray-2.4 ray-team-created Ray Team created P1 Issue that should be fixed within a few weeks and removed P2 Important issue, but not time-critical labels Feb 17, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
docs An issue or change related to documentation P1 Issue that should be fixed within a few weeks Ray-2.4 ray-team-created Ray Team created tune Tune-related issues
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants