`trw.train.compatibility`¶

Module Contents¶

Classes¶

SwishCompat

For compatibility with old PyTorch versions

Functions¶

`affine_grid`(theta: torch.Tensor, size: List[int], align_corners: Optional[bool]) → torch.Tensor	Compatibility layer for new arguments introduced in pytorch 1.3
`grid_sample`(input: torch.Tensor, grid: torch.Tensor, mode: str = 'bilinear', padding_mode: str = 'zeros', align_corners: bool = None) → torch.Tensor	Compatibility layer for argument change between pytorch <= 1.2 and pytorch > 1.3

Attributes¶

`torch_version`
`Swish`
`inv`
`torch_linalg_norm`
`Identity`

trw.train.compatibility.torch_version¶

trw.train.compatibility.affine_grid(theta: torch.Tensor, size: List[int], align_corners: Optional[bool]) → torch.Tensor¶

Compatibility layer for new arguments introduced in pytorch 1.3

See torch.nn.functional.affine_grid()

trw.train.compatibility.grid_sample(input: torch.Tensor, grid: torch.Tensor, mode: str = 'bilinear', padding_mode: str = 'zeros', align_corners: bool = None) → torch.Tensor¶

Compatibility layer for argument change between pytorch <= 1.2 and pytorch > 1.3

See torch.nn.functional.grid_sample()

class trw.train.compatibility.SwishCompat¶

Bases: torch.nn.Module