[Unified Checkpoint] update merge tensor parallel #8856
Merged
Conversation
This file contains 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
|
Thanks for your contribution! |
ZHUI
reviewed
Aug 1, 2024
| ret = distributed_allgather(tensor, group=tp_group, offload=False) | ||
| # Get tensor size | ||
| tensor_bytes = tensor.numel().item() * dtype_byte_size(tensor.dtype) * tp_group.nranks | ||
| if tensor_bytes >= 5368709120: # temporarily set 5GB as threshold |
There was a problem hiding this comment.
No Magic Number。 tensor_bytes > 5*1024*1024*1024
建议再设置小一点。
DesmonDay
force-pushed
the
update_uc_merge
branch
from
0a21315 to
8ddf398
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #8856 +/- ##
===========================================
- Coverage 55.43% 55.42% -0.02%
===========================================
Files 631 631
Lines 98544 98577 +33
===========================================
+ Hits 54632 54633 +1
- Misses 43912 43944 +32 ☔ View full report in Codecov by Sentry. |
DrownFish19
pushed a commit
to DrownFish19/PaddleNLP
that referenced
this pull request
Aug 2, 2024
DrownFish19
pushed a commit
to DrownFish19/PaddleNLP
that referenced
this pull request
Aug 2, 2024
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Outdated suggestions cannot be applied.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.
PR types
Others
PR changes
Others
Description
When merging large tensors, there may be not enough GPU memory to allocate due to VRAM fragmentation. Therefore we split the large tensors into smaller parts, and then merge these tensors on CPU.