fix a bug on torch_dtype argument in from_single_file of ControlNetModel

[xuyxu]

The following code snippet will throw TypeError for now:

import torch
from diffusers import ControlNetModel
controlnet = ControlNetModel.from_single_file(
    "./control_sd15_canny.pth", torch_dtype=torch.float16, local_files_only=True
)

which gives us:

Traceback (most recent call last):
  File "D:\AI\diffusers_test.py", line 5, in <module>
    controlnet = ControlNetModel.from_single_file(
  File "D:\Programming\miniconda3\lib\site-packages\diffusers\loaders.py", line 2665, in from_single_file
    controlnet.to(torch_dtype=torch_dtype)
  File "D:\Programming\miniconda3\lib\site-packages\torch\nn\modules\module.py", line 1141, in to
    device, dtype, non_blocking, convert_to_format = torch._C._nn._parse_to(*args, **kwargs)
TypeError: to() received an invalid combination of arguments - got (torch_dtype=torch.dtype, ), but expected one of:
 * (torch.device device, torch.dtype dtype, bool non_blocking, bool copy, *, torch.memory_format memory_format)
 * (torch.dtype dtype, bool non_blocking, bool copy, *, torch.memory_format memory_format)
 * (Tensor tensor, bool non_blocking, bool copy, *, torch.memory_format memory_format)

The reason is that the parameter name torch_dtype here is wrong, it should be dtype instead of torch_dtype, according to the function signature of the inherited to method from Pytorch module.py.

[patrickvonplaten]

Cool!