[WIP][Feature][Bugfix] WandB setup/finish simplified + callback. - #381
Closed
sami-bg wants to merge 2 commits into
Closed
[WIP][Feature][Bugfix] WandB setup/finish simplified + callback.#381sami-bg wants to merge 2 commits into
sami-bg wants to merge 2 commits into
Conversation
Contributor
|
Nice, thanks @sami-bg ! did you try running with DDP and preemption/requeue to make sure everything is safely dumped/loaded re-init every time? |
Contributor
Author
|
Not yet! Sorry forgot to add [WIP]. Will test this soon |
Contributor
|
Hi @sami-bg ! I did an update on wandb callbacks in the last few merge, could you please double check and if spt now has all the functionalities you were implementing, please close this PR, thank you! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Extracts all wandb lifecycle logic (init, config sync, offline data dump, wandb.finish()) from Manager into a new WandbCallback that uses Lightning's setup()/teardown() hooks, so wandb.finish() is called automatically after each trainer stage. This fixes run leaks in multirun sweeps (fit then test) without requiring manual cleanup
Moves WandbCheckpoint out of checkpoint_sklearn.py into wandb_lifecycle.py alongside WandbCallback, and fixes its docstring which was copy-pasted from the sklearn checkpoint
Adds unit tests covering config flattening, offline file I/O, SLURM requeue directory discovery, setup/teardown state management, on_exception crash recovery, and checkpoint run-ID continuity
Checklist