Skip to content

[Bugfix] Fix torch.compile x LoRA for PyTorch 2.8 #20823

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
merged 1 commit into from
Jul 12, 2025
Merged

[Bugfix] Fix torch.compile x LoRA for PyTorch 2.8 #20823

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
14 changes: 8 additions & 6 deletions vllm/lora/layers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -240,17 +240,19 @@ def set_lora(
def forward(self, x: torch.Tensor) -> torch.Tensor:
added_tokens_mask = torch.where(x > self.base_layer.org_vocab_size - 1,
1, 0)
embeddings_indices = torch.narrow(
self.punica_wrapper._embeddings_indices, 1, 0, x.size(0))

indices = embeddings_indices[1]
# NB: Don't use torch.narrow here. torch.narrow triggers some
# Dynamic Shape specialization in torch.compile
num_tokens = x.shape[0]
indices_1 = self.punica_wrapper._embeddings_indices[1][:num_tokens]
indices_0 = self.punica_wrapper._embeddings_indices[0][:num_tokens]

full_lora_a_embeddings = F.embedding(
x + indices,
x + indices_1,
self.lora_a_stacked_2d,
)
indices = embeddings_indices[0]
full_output = self.base_layer.forward(x +
(indices * added_tokens_mask))
(indices_0 * added_tokens_mask))

full_output_org = full_output
if full_output.ndim == 3:
Expand Down