Temporary workaround for loading best model at end with DeepSpeed#95
Merged
Conversation
|
The documentation is not available anymore as the PR was closed or merged. |
libinta
approved these changes
Sep 9, 2022
hsubramony
pushed a commit
that referenced
this pull request
Mar 13, 2024
yeonsily
pushed a commit
that referenced
this pull request
Mar 22, 2024
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
3 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Loading the best model at the end of training with
--load_best_model_at_endfails with the current version of Habana DeepSpeed (0.6.1, see huggingface/transformers#17114).This PR brings a temporary workaround where the best model at the end of training is loaded as a regular PyTorch model and not as a DeepSpeed engine. This should not be an issue since the best model is loaded for inference only and ZeRO-3 has not been validated yet (see here) while ZeRO-1/2 are useful for training only.
Before submitting