[WIP] fix transforrmers api change at 5.2.0#1647
Open
UbeCc wants to merge 10 commits intoTHUDM:mainfrom
Open
Conversation
UbeCc
changed the title
zhuzilin
approved these changes
Feb 28, 2026
- Create tensorboard_dir if it doesn't exist before passing to trace handler - Create memory_snapshot_dir before profiler tries to save snapshots - Prevents FileNotFoundError when profiling with missing directories Fixes issue where OOM observer fails when memory_snapshot_dir doesn't exist. Made-with: Cursor
…snapshot saving - Add try-except blocks in OOM observer and stop() method with detailed logging - Verify that snapshot files are actually created and log file sizes - Convert Path objects to strings explicitly for better compatibility - Add stderr tracebacks for better error visibility - Log snapshot path before attempting to save for debugging - Use os.makedirs instead of Path.mkdir for more robust directory creation This helps diagnose issues where snapshots appear to be saved but files are empty. Made-with: Cursor
… actors - Convert memory_snapshot_dir and tensorboard_dir to absolute paths using .resolve() - Prevents issues where Ray actors save files in different working directories - Add log message showing absolute path being used - Convert Path objects to strings for memray compatibility Fixes issue where snapshot files appear to be saved but end up in wrong location when training is distributed across Ray actors with different working directories. Made-with: Cursor
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
No description provided.