Skip to content

feat: support DTensor CP in DPO and SFT #798

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
terrykong merged 3 commits into from
Aug 7, 2025
Merged

feat: support DTensor CP in DPO and SFT #798

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f3ed48c
support DTensor + CP in DPO and SFT
ashors1 Jul 30, 2025
d8c20c7
add assertion to loss functions
ashors1 Jul 30, 2025
313c036
Revert "add assertion to loss functions"
ashors1 Aug 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions nemo_rl/algorithms/loss_functions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -325,6 +325,7 @@ def __call__(
token_mask = data["token_mask"][:, 1:]
sample_mask = data["sample_mask"]
mask = token_mask * sample_mask.unsqueeze(-1)
seq_index = data.get("seq_index", None)
Comment thread
ashors1 marked this conversation as resolved.

next_token_logits = next_token_logits.to(torch.float32)

Expand All @@ -346,7 +347,7 @@ def __call__(
token_logprobs = token_logprobs[:, : data["input_ids"].shape[1] - 1]
elif isinstance(next_token_logits, torch.distributed.tensor.DTensor):
token_logprobs = get_logprobs_from_vocab_parallel_logits(
next_token_logits, data["input_ids"]
next_token_logits, data["input_ids"], seq_index=seq_index
)
else:
next_tokens = data["input_ids"][:, 1:].cuda() # Skip first token
Expand Down Expand Up @@ -580,6 +581,7 @@ def _dpo_loss(
## TODO(@ashors): there's some duplicate code here with the NLLLoss function. We should refactor
token_mask = data["token_mask"][:, 1:]
sample_mask = data["sample_mask"]
seq_index = data.get("seq_index", None)

next_token_logits = next_token_logits.to(torch.float32)
if vocab_parallel_group is not None:
Expand All @@ -599,7 +601,7 @@ def _dpo_loss(
token_logprobs = token_logprobs[:, : data["input_ids"].shape[1] - 1]
elif isinstance(next_token_logits, torch.distributed.tensor.DTensor):
token_logprobs = get_logprobs_from_vocab_parallel_logits(
next_token_logits, data["input_ids"]
next_token_logits, data["input_ids"], seq_index=seq_index
)
else:
next_tokens = data["input_ids"][:, 1:].cuda() # Skip first token
Expand Down