-
Notifications
You must be signed in to change notification settings - Fork 25.2k
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
feature: update wandb callback to upload checkpoints #21035
feature: update wandb callback to upload checkpoints #21035
Conversation
- removes duplicated run names from wandb workspace - models can be logged with corresponding run names
The documentation is not available anymore as the PR was closed or merged. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for your PR! I left a couple of comments.
Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>
Hey, @sgugger : Thanks for the quick review and suggestions. I've resolved all the issues. 🤗 |
@parambharat There is a weird issue with the tests. Could you try refreshing your permissions as shown here and then pushing an empty commit? |
Hey, @sgugger. I ran the fix that you suggested and all checks pass now. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for updating! I left two suggestions on formatting :)
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
Thank you @stevhliu. I've committed your recommendations. |
Thanks again for your contribution! |
* docs: add wandb metrics and model checkpointing to callback docstrings * docs: update reference to wandb documentation * fix: change default of `"WANDB_WATCH"` from ``"gradients"` to ``"false"` * feature: add `on_save` method and update `"WANDB_LOG_MODEL` behaviour * fix: use default wandb run names instead of `output_dir` - removes duplicated run names from wandb workspace - models can be logged with corresponding run names * fix: edit deprecation warning based on review suggestions Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com> * fix: change indentation of docstrings * fix: change indentation of docstrings and run fixup * fix: empty commit for circleci permissions issue * fix: format deprecation doc strings review suggestion Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * docs: Highlight WANDB_DISABLED arg in documentaion Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * fix: run fixup after updating docstrings Co-authored-by: Bharat Ramanathan <ramanathan.parameshwaran@gohuddl.com> Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com> Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* docs: add wandb metrics and model checkpointing to callback docstrings * docs: update reference to wandb documentation * fix: change default of `"WANDB_WATCH"` from ``"gradients"` to ``"false"` * feature: add `on_save` method and update `"WANDB_LOG_MODEL` behaviour * fix: use default wandb run names instead of `output_dir` - removes duplicated run names from wandb workspace - models can be logged with corresponding run names * fix: edit deprecation warning based on review suggestions Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com> * fix: change indentation of docstrings * fix: change indentation of docstrings and run fixup * fix: empty commit for circleci permissions issue * fix: format deprecation doc strings review suggestion Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * docs: Highlight WANDB_DISABLED arg in documentaion Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * fix: run fixup after updating docstrings Co-authored-by: Bharat Ramanathan <ramanathan.parameshwaran@gohuddl.com> Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com> Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
What does this PR do?
The PR updates the
WandbCallback
with the following changes:on_save
method to upload model checkpoints as artifacts.WANDB_WATCH
fromgradients
tofalse
. This enables quicker training when defaults are used. The user can easily change this behavior by setting the env variable.WANDB_LOG_MODEL
variable frombool
tostr
allowing for different settings to upload artifacts.run_name
fromoutput_dir
to wandb auto generated name. this avoids duplication of run names in wandb workspaceBefore submitting
Pull Request section?
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Examples