Skip to content
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.

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

bugfix for npu not support float64 #10123

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
5 changes: 3 additions & 2 deletions examples/community/fresco_v2v.py
Original file line number Diff line number Diff line change
Expand Up @@ -404,10 +404,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions examples/community/matryoshka.py
Original file line number Diff line number Diff line change
Expand Up @@ -2806,10 +2806,11 @@ def get_time_embed(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -1031,10 +1031,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor([current_timestep], dtype=dtype, device=latent_model_input.device)
elif len(current_timestep.shape) == 0:
current_timestep = current_timestep[None].to(latent_model_input.device)
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -258,10 +258,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/controlnets/controlnet.py
Original file line number Diff line number Diff line change
Expand Up @@ -740,10 +740,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/controlnets/controlnet_sparsectrl.py
Original file line number Diff line number Diff line change
Expand Up @@ -671,10 +671,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/controlnets/controlnet_union.py
Original file line number Diff line number Diff line change
Expand Up @@ -681,10 +681,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/controlnets/controlnet_xs.py
Original file line number Diff line number Diff line change
Expand Up @@ -1088,10 +1088,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/unets/unet_2d_condition.py
Original file line number Diff line number Diff line change
Expand Up @@ -915,10 +915,11 @@ def get_time_embed(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/unets/unet_3d_condition.py
Original file line number Diff line number Diff line change
Expand Up @@ -624,10 +624,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/unets/unet_i2vgen_xl.py
Original file line number Diff line number Diff line change
Expand Up @@ -575,10 +575,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass `timesteps` as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timesteps, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/unets/unet_motion_model.py
Original file line number Diff line number Diff line change
Expand Up @@ -2114,10 +2114,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/models/unets/unet_spatio_temporal_condition.py
Original file line number Diff line number Diff line change
Expand Up @@ -402,10 +402,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/audioldm2/modeling_audioldm2.py
Original file line number Diff line number Diff line change
Expand Up @@ -768,10 +768,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -1163,10 +1163,11 @@ def forward(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = sample.device.type == "mps"
is_npu = sample.device.type == "npu"
if isinstance(timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=sample.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(sample.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/dit/pipeline_dit.py
Original file line number Diff line number Diff line change
Expand Up @@ -187,10 +187,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(timesteps, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
timesteps = torch.tensor([timesteps], dtype=dtype, device=latent_model_input.device)
elif len(timesteps.shape) == 0:
timesteps = timesteps[None].to(latent_model_input.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/latte/pipeline_latte.py
Original file line number Diff line number Diff line change
Expand Up @@ -797,10 +797,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor([current_timestep], dtype=dtype, device=latent_model_input.device)
elif len(current_timestep.shape) == 0:
current_timestep = current_timestep[None].to(latent_model_input.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/lumina/pipeline_lumina.py
Original file line number Diff line number Diff line change
Expand Up @@ -806,10 +806,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor(
[current_timestep],
dtype=dtype,
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/pag/pipeline_pag_pixart_sigma.py
Original file line number Diff line number Diff line change
Expand Up @@ -807,10 +807,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor([current_timestep], dtype=dtype, device=latent_model_input.device)
elif len(current_timestep.shape) == 0:
current_timestep = current_timestep[None].to(latent_model_input.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/pixart_alpha/pipeline_pixart_alpha.py
Original file line number Diff line number Diff line change
Expand Up @@ -907,10 +907,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor([current_timestep], dtype=dtype, device=latent_model_input.device)
elif len(current_timestep.shape) == 0:
current_timestep = current_timestep[None].to(latent_model_input.device)
Expand Down
5 changes: 3 additions & 2 deletions src/diffusers/pipelines/pixart_alpha/pipeline_pixart_sigma.py
Original file line number Diff line number Diff line change
Expand Up @@ -822,10 +822,11 @@ def __call__(
# TODO: this requires sync between CPU and GPU. So try to pass timesteps as tensors if you can
# This would be a good case for the `match` statement (Python 3.10+)
is_mps = latent_model_input.device.type == "mps"
is_npu = latent_model_input.device.type == "npu"
if isinstance(current_timestep, float):
dtype = torch.float32 if is_mps else torch.float64
dtype = torch.float32 if (is_mps or is_npu) else torch.float64
else:
dtype = torch.int32 if is_mps else torch.int64
dtype = torch.int32 if (is_mps or is_npu) else torch.int64
current_timestep = torch.tensor([current_timestep], dtype=dtype, device=latent_model_input.device)
elif len(current_timestep.shape) == 0:
current_timestep = current_timestep[None].to(latent_model_input.device)
Expand Down
Loading