Skip to content
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

Integrate Orbax's emergency checkpoint. #820

Open
wants to merge 8 commits into
base: main
Choose a base branch
from
+1,126 −10
Open

Integrate Orbax's emergency checkpoint. #820

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
cbaabdc
Integrate Orbax emergency checkpoint
hanzhi713 Jan 8, 2025
9d86aaf
Address comments
hanzhi713 Jan 16, 2025
16d877a
comment
hanzhi713 Jan 16, 2025
c1a476d
Address comments
hanzhi713 Jan 27, 2025
140a18f
Upgrade orbax
hanzhi713 Jan 30, 2025
1af8ce7
Improve comments
hanzhi713 Jan 30, 2025
4fc557a
Improve comments
hanzhi713 Jan 31, 2025
fb7a4a3
Update for new orbax versions
hanzhi713 Feb 1, 2025
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
5 changes: 2 additions & 3 deletions axlearn/common/checkpointer_orbax.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -346,9 +346,8 @@ def _restore_args(x: Any) -> ocp.RestoreArgs:
state=ocp.args.PyTreeRestore(item=state, restore_args=restore_args),
),
)
except TypeError as e:
# Orbax hits TypeError if there are no checkpoints, since it attempts to format `None`
# as the step dir.
except FileNotFoundError as e:
# Orbax raises FileNotFoundError if there are no checkpoints.
if step is not None:
raise ValueError(f"Failed to restore at step {step}.") from e
logging.info("Could not find any completed checkpoints under %s: %s", cfg.dir, e)
Expand Down
Loading