[demo] verify fully_shard([norm, head]) and fully_shard([tok_embedding, norm, head]) works with chunked loss#2976
Closed
weifengpy wants to merge 6 commits intopytorch:mainfrom
Closed
[demo] verify fully_shard([norm, head]) and fully_shard([tok_embedding, norm, head]) works with chunked loss#2976weifengpy wants to merge 6 commits intopytorch:mainfrom
weifengpy wants to merge 6 commits intopytorch:mainfrom
Conversation
Implements chunked cross-entropy loss that splits the sequence dimension into N chunks, computing lm_head projection and CE loss per-chunk to avoid materializing the full [B, L, V] logits tensor at once. Key components: - ChunkedCELoss: wraps lm_head + ce_loss with chunked forward/backward - GradAccumulator: pre-allocated buffer for assembling chunk gradients - _no_reshard_after_backward: FSDP2 context to avoid N all-gathers - skip_lm_head kwarg on Decoder.forward() for the detach boundary - ChunkedCELossFactory: deferred initialization (model not available at build time) - Trainer integration with dedicated forward_backward_step branch
…CELoss - Add loss_num_chunks to TrainingConfig (default 1, no-op) - Trainer auto-wraps loss_fn in ChunkedCELossFactory when loss_num_chunks > 1 - Integration tests for FSDP, FSDP+TP(SP), FSDP+CP, FSDP+TP+CP, FSDP+compile
FSDP2's backward hooks are one-shot per forward pass. The previous approach of calling self.lm_head(h_chunk) triggered FSDP2's backward hooks during chunk backward, leaving no hooks for the decoder backward (h.backward(grad)), causing zero gradients on model parameters. Fix: Use F.linear(h_chunk, lm_weight) to bypass FSDP2 module hooks during chunk computation. Use (h * accumulated_grad).sum().backward() instead of h.backward(grad) to properly trigger FSDP2's hooks in a single backward pass.
Replace bare function + build_fn pattern with proper loss classes. CrossEntropyLoss and MSELoss encapsulate compilation logic internally. The old function names (cross_entropy_loss, mse_loss) remain as public API for backward compatibility. build_cross_entropy_loss and build_mse_loss now return class instances.
meta-cla
Bot
added
the
CLA Signed
weifengpy
changed the title
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.
remove if-else base on chunked loss in apply_fsdp: #2937