Add hysteresis support for AMP gradient scale update#1733
Merged
crcrpar merged 8 commits intoNVIDIA:masterfrom Sep 30, 2023
Merged
Add hysteresis support for AMP gradient scale update#1733crcrpar merged 8 commits intoNVIDIA:masterfrom
crcrpar merged 8 commits intoNVIDIA:masterfrom
Conversation
…s by tricking torch autograd (NVIDIA#1715) * input grad checks out * adding clamp gamma * Both old and proposed implementation checks out * 2 tests not yet passed due to numerical issues * mem_eff works * fast-layer-norm done * Moving mem-eff to templates * Relax tolerance for memory efficient backward * Fix backward api of python
* Add distopt support for param syncs with non-floating-point dtypes Signed-off-by: Tim Moon <tmoon@nvidia.com> * Update apex/contrib/optimizers/distributed_fused_adam.py Co-authored-by: Masaki Kozuki <mkozuki@nvidia.com> --------- Signed-off-by: Tim Moon <tmoon@nvidia.com> Co-authored-by: Masaki Kozuki <mkozuki@nvidia.com>
minitu
force-pushed
the
update_scale
branch
from
c4dc727 to
2657766
Compare
crcrpar
reviewed
Sep 30, 2023
Contributor
Author
|
@crcrpar Thanks, addressed your comments |
crcrpar
approved these changes
Sep 30, 2023
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
4 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR adds support for hysteresis in AMP gradient scale update.