Skip to content

Remove outdated warning about loading full-state checkpoints in FSDP #18208

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.

Sign up for GitHub

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

Merged
merged 1 commit into from
Aug 1, 2023
Merged

Remove outdated warning about loading full-state checkpoints in FSDP #18208

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 0 additions & 6 deletions docs/source-fabric/advanced/model_parallel/fsdp.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -387,12 +387,6 @@ You can easily load checkpoints saved by Fabric to resume training:

Fabric will automatically recognize whether the provided path contains a checkpoint saved with ``state_dict_type="full"`` or ``state_dict_type="sharded"``.

.. warning::

Loading a full-state checkpoint will replicate the file in CPU RAM for every GPU.
For very large checkpoints/models, you may run out of memory and your program will crash.
If this happens, save using the “sharded” checkpoint format instead (default).


----

Expand Down