Switches BBS example to use mbsize=3 and gas=2 to fit in 16GB of memory. #341
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
ShadenSmith
requested review from
arashashari,
awan-10,
cli99,
conglongli,
eltonzheng,
jeffra,
minjiaz,
niumanar,
RezaYazdaniAminabadi,
samyam and
tjruwase
as code owners
tjruwase
approved these changes
Sep 1, 2020
|
What issue is this PR related to? |
|
We are ramping up additional testing machines with 16GB GPUs. BBS is the only test that requires >16GB of memory to complete. As an added bonus this now adds model test coverage for ZeRO-2 with gradient accumulation too :-). Do you see any issues with decreasing the microbatch size for these model tests? The effective batch size is untouched. |
|
Thanks for the clarification. I don't see any issues with decreasing micro batch size. I was just curious about the motivation. |
|
Sure sure, was just wondering too :-). I haven't worked with this test much in the past. |
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Outdated suggestions cannot be applied.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.
No description provided.