adding TiledFusedLogitsLoss #7437
Merged
Conversation
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
Signed-off-by: Stas Bekman <[email protected]>
Signed-off-by: Stas Bekman <[email protected]>
Signed-off-by: Stas Bekman <[email protected]>
Signed-off-by: Stas Bekman <[email protected]>
Collaborator
Author
|
@tjruwase, could you please review? it should be easy to review since it's a fork of an existing |
sfc-gh-truwase
approved these changes
Jul 30, 2025
| with torch.enable_grad(): | ||
| args = (self, x_shard, y_shard) | ||
| if mask is not None: | ||
| args.append(mask_shards[i]) |
qimcis
pushed a commit
to qimcis/DeepSpeed
that referenced
this pull request
Jul 31, 2025
This PR adds `TiledFusedLogitsLoss` for an efficient fused logits+loss computation - this version pre-calculates grads in `forward`, avoiding recomputation in the backward (similar to the Liger-Kernel implementation). --------- Signed-off-by: Stas Bekman <[email protected]> Co-authored-by: Aurick Qiao <[email protected]> Signed-off-by: qimcis <[email protected]>
LYMDLUT
pushed a commit
to LYMDLUT/DeepSpeed
that referenced
this pull request
Aug 20, 2025
This PR adds `TiledFusedLogitsLoss` for an efficient fused logits+loss computation - this version pre-calculates grads in `forward`, avoiding recomputation in the backward (similar to the Liger-Kernel implementation). --------- Signed-off-by: Stas Bekman <[email protected]> Co-authored-by: Aurick Qiao <[email protected]> Signed-off-by: lym <[email protected]>
mauryaavinash95
pushed a commit
to DataStates/DeepSpeed
that referenced
this pull request
Oct 4, 2025
This PR adds `TiledFusedLogitsLoss` for an efficient fused logits+loss computation - this version pre-calculates grads in `forward`, avoiding recomputation in the backward (similar to the Liger-Kernel implementation). --------- Signed-off-by: Stas Bekman <[email protected]> Co-authored-by: Aurick Qiao <[email protected]>
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.
This PR adds
TiledFusedLogitsLossfor an efficient fused logits+loss computation - this version pre-calculates grads inforward, avoiding recomputation in the backward (similar to the Liger-Kernel implementation).