You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 3, 2023. It is now read-only.
As the title already suggests.
I was looking into the latest examples of PyTorch with Ray Tune here and found that resuming from a checkpoint and reporting to tune is now handled via the session API provided by Ray AIR.
Is it necessary to adapt TuneReportCallback and TuneReportCheckpointCallback accordingly?
The text was updated successfully, but these errors were encountered:
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
As the title already suggests.
I was looking into the latest examples of PyTorch with Ray Tune here and found that resuming from a checkpoint and reporting to tune is now handled via the session API provided by Ray AIR.
Is it necessary to adapt
TuneReportCallback
andTuneReportCheckpointCallback
accordingly?The text was updated successfully, but these errors were encountered: