Skip to content

Support RotaryEmbedding in GQA for webgpu ep #1847

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
kunal-vaishnavi merged 1 commit into from
Oct 30, 2025
Merged

Support RotaryEmbedding in GQA for webgpu ep #1847

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
4 changes: 2 additions & 2 deletions src/python/py/models/builder.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -369,14 +369,14 @@ def make_attention_init(self):
# Some EPs don't support packed Q/K/V for GQA yet
# Packed MatMul with LoRA/QLoRA is not currently supported
self.attention_attrs["use_packed_matmul"] = (
self.ep not in ["dml", "webgpu"]
self.ep not in ["dml"]
and not self.matmul_attrs["use_lora"]
and not self.attention_attrs["q_norm"]
and not self.attention_attrs["k_norm"]
)

# Some EPs don't support fusing rotary embeddings inside GQA yet
self.attention_attrs["use_rope_in_attn"] = self.ep not in ["dml", "webgpu"]
self.attention_attrs["use_rope_in_attn"] = self.ep not in ["dml"]
if self.attention_attrs["use_rope_in_attn"]:
# GQA + Rot.Emb. does not require `position_ids` as input
self.input_names.remove("position_ids")
Expand Down
Loading