Move loss_fn into model definition in PP example#324
Merged
Conversation
xmfan
added a commit
that referenced
this pull request
Feb 24, 2026
Instead of passing loss_fn to AutoParallelPP, wrap the model in a ModelWithLoss module that bakes cross-entropy loss into forward(). This makes the example compatible with tracing self.model directly. stack-info: PR: #324, branch: xmfan/stack/25
xmfan
force-pushed
the
xmfan/stack/25
branch
from
c933640 to
f208859
Compare
meta-cla
bot
added
the
CLA Signed
fmassa
approved these changes
Feb 24, 2026
Contributor
fmassa
left a comment
fmassa
left a comment
There was a problem hiding this comment.
LGTM, thanks! Then I suppose we can remove the custom export that we have?
Instead of passing loss_fn to AutoParallelPP, wrap the model in a ModelWithLoss module that bakes cross-entropy loss into forward(). This makes the example compatible with tracing self.model directly. stack-info: PR: #324, branch: xmfan/stack/25
xmfan
force-pushed
the
xmfan/stack/25
branch
from
f208859 to
b97e75a
Compare
Member
Author
|
@fmassa you mean clean up AutoParallelPP? or something else? |
xmfan
added a commit
that referenced
this pull request
Mar 2, 2026
Same approach as #324 for example_pp_graph_passes.py: wrap the last stage in a ModelWithLoss module instead of passing loss_fn to AutoParallelPP, which no longer supports it. Authored with Claude.
xmfan
added a commit
that referenced
this pull request
Mar 4, 2026
Same approach as #324 for example_pp_graph_passes.py: wrap the last stage in a ModelWithLoss module instead of passing loss_fn to AutoParallelPP, which no longer supports it. Authored with Claude.
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
2 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.
Stacked PRs:
Move loss_fn into model definition in PP example
Instead of passing loss_fn to AutoParallelPP, wrap the model in a
ModelWithLoss module that bakes cross-entropy loss into forward().
This makes the example compatible with tracing self.model directly.