wraps DDP models with DSD #857
Open
Conversation
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
|
This pull request was exported from Phabricator. Differential Revision: D59234083 |
Summary: Pull Request resolved: #857 Distributed State Dict is the current suggested way from PyTorch for ensuring parallelized models state dicts are compatible with save/loads in Single process or re-sharding scenarios. This diff updates dcp_saver to use DSD for DDP models. A good idea would be wrap all models in TNT with DSD, as this could replace some of the wrapper logic for FSDP and would guarantee future compat. N5551629 also contains a workaround for current DDP model saved before this diff, by manually removing the "module." prefix in the checkpoint. Differential Revision: D59234083
LucasLLC
force-pushed
the
export-D59234083
branch
from
5818bb8 to
435b1cb
Compare
|
This pull request was exported from Phabricator. Differential Revision: D59234083 |
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Outdated suggestions cannot be applied.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.
Summary:
Distributed State Dict is the current suggested way from PyTorch for ensuring parallelized models state dicts are compatible with save/loads in Single process or re-sharding scenarios.
This diff updates dcp_saver to use DSD for DDP models. A good idea would be wrap all models in TNT with DSD, as this could replace some of the wrapper logic for FSDP and would guarantee future compat.
N5551629 also contains a workaround for current DDP model saved before this diff, by manually removing the "module." prefix in the checkpoint.
Differential Revision: D59234083