[🐛 BUG] Fix bugs when collecting results from mp.spawn
in multi-GPU training
#1875
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Outdated suggestions cannot be applied.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.
Bug description
We are unable to collect results directly from
mp.spawn
, e.g.,RecBole/run_recbole_group.py
Lines 25 to 38 in 96eb311
torch.multiprocessing.Queue can be applied for this purpose.
Changelog in this PR
mp.spawn
;distributed training
;