Skip to content

Use float32 on mps or npu in transformer_hidream_image's rope #11316

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 2 commits into from
Apr 14, 2025
Merged

Use float32 on mps or npu in transformer_hidream_image's rope #11316

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7c508d4
Use float32 on mps or npu in transformer_hidream_image's rope
hlky Apr 14, 2025
89158ce
Merge branch 'main' into mps-hidream
hlky Apr 14, 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
7 changes: 6 additions & 1 deletion src/diffusers/models/transformers/transformer_hidream_image.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,7 +95,12 @@ def forward(self, latent):
def rope(pos: torch.Tensor, dim: int, theta: int) -> torch.Tensor:
assert dim % 2 == 0, "The dimension must be even."

scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
is_mps = pos.device.type == "mps"
is_npu = pos.device.type == "npu"

dtype = torch.float32 if (is_mps or is_npu) else torch.float64

scale = torch.arange(0, dim, 2, dtype=dtype, device=pos.device) / dim
omega = 1.0 / (theta**scale)

batch_size, seq_length = pos.shape
Expand Down
Loading