API for obtaining global gradient norm#1292
Merged
tjruwase merged 4 commits intobig-sciencefrom Aug 9, 2021
Merged
Conversation
Collaborator
|
Thank you for working on this, @tjruwase and then we will need Shaden's deepspeedai/Megatron-DeepSpeed#8 on the Megatron side once this is merged. |
ShadenSmith
approved these changes
Aug 9, 2021
Contributor
ShadenSmith
left a comment
There was a problem hiding this comment.
@samyam may also want to take a look
jeffra
pushed a commit
that referenced
this pull request
Sep 8, 2021
* FP16 fused and unfused grad norm query. * API for obtaining global unclipped gradient norm across parameter groups * Use global norm not group norms Co-authored-by: Shaden Smith <shaden.smith@microsoft.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
API for obtaining global unclipped gradient norm across all parameters groups. Based off #1286.
Optimizers are solely responsible for computing gradient norms. Gradient norms are computed (or refreshed) in optimizer.step().
@stas00 FYI