Mirrors
/
stable-diffusion-webui
mirror of https://github.com/AUTOMATIC1111/stable-diffusion-webui.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix cumsum for MPS in newer torch 

The prior fix assumed that testing int16 was enough to determine if a fix is needed, but a recent fix for cumsum has int16 working but not bool.
This commit is contained in:
brkirch 2023-01-12 08:00:38 -05:00
parent 0b8911d883
commit a255dac4f8
1 changed files with 7 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
modules/devices.py
View File

@ -139,8 +139,10 @@ orig_Tensor_cumsum = torch.Tensor.cumsum
def cumsum_fix(input, cumsum_func, *args, **kwargs):
if input.device.type == 'mps':
output_dtype = kwargs.get('dtype', input.dtype)
if any(output_dtype == broken_dtype for broken_dtype in [torch.bool, torch.int8, torch.int16, torch.int64]):
if output_dtype == torch.int64:
return cumsum_func(input.cpu(), *args, **kwargs).to(input.device)
elif cumsum_needs_bool_fix and output_dtype == torch.bool or cumsum_needs_int_fix and (output_dtype == torch.int8 or output_dtype == torch.int16):
return cumsum_func(input.to(torch.int32), *args, **kwargs).to(torch.int64)
return cumsum_func(input, *args, **kwargs)
@ -151,8 +153,9 @@ if has_mps():
torch.nn.functional.layer_norm = layer_norm_fix
torch.Tensor.numpy = numpy_fix
elif version.parse(torch.__version__) > version.parse("1.13.1"):
if not torch.Tensor([1,2]).to(torch.device("mps")).equal(torch.Tensor([1,1]).to(torch.device("mps")).cumsum(0, dtype=torch.int16)):
torch.cumsum = lambda input, *args, **kwargs: ( cumsum_fix(input, orig_cumsum, *args, **kwargs) )
torch.Tensor.cumsum = lambda self, *args, **kwargs: ( cumsum_fix(self, orig_Tensor_cumsum, *args, **kwargs) )
cumsum_needs_int_fix = not torch.Tensor([1,2]).to(torch.device("mps")).equal(torch.ShortTensor([1,1]).to(torch.device("mps")).cumsum(0))
cumsum_needs_bool_fix = not torch.BoolTensor([True,True]).to(device=torch.device("mps"), dtype=torch.int64).equal(torch.BoolTensor([True,False]).to(torch.device("mps")).cumsum(0))
torch.cumsum = lambda input, *args, **kwargs: ( cumsum_fix(input, orig_cumsum, *args, **kwargs) )
torch.Tensor.cumsum = lambda self, *args, **kwargs: ( cumsum_fix(self, orig_Tensor_cumsum, *args, **kwargs) )
orig_narrow = torch.narrow
torch.narrow = lambda *args, **kwargs: ( orig_narrow(*args, **kwargs).clone() )