Skip to content

Update Mish to default PyTorch 1.9 version #2733

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
wyli merged 6 commits into from
Aug 11, 2021
Merged

Update Mish to default PyTorch 1.9 version #2733

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
4d29fa6
Update Mish to default PyTorch 1.9 version
digantamisra98 Aug 9, 2021
6b07bf4
Update activation.py
digantamisra98 Aug 9, 2021
f275df7
update based on comments
wyli Aug 10, 2021
5384319
impl. with optional import
wyli Aug 10, 2021
6569e73
fixes mypy error
wyli Aug 11, 2021
5c1a2ab
Merge branch 'dev' into patch-1
wyli Aug 11, 2021
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
23 changes: 21 additions & 2 deletions monai/networks/blocks/activation.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,19 @@
import torch
from torch import nn

from monai.utils import optional_import

if optional_import("torch.nn.functional", name="mish")[1]:

def monai_mish(x):
return torch.nn.functional.mish(x, inplace=True)


else:

def monai_mish(x):
return x * torch.tanh(torch.nn.functional.softplus(x))


class Swish(nn.Module):
r"""Applies the element-wise function:
Expand All @@ -30,6 +43,8 @@ class Swish(nn.Module):

Examples::

>>> import torch
>>> from monai.networks.layers.factories import Act
>>> m = Act['swish']()
>>> input = torch.randn(2)
>>> output = m(input)
Expand Down Expand Up @@ -85,6 +100,8 @@ class MemoryEfficientSwish(nn.Module):

Examples::

>>> import torch
>>> from monai.networks.layers.factories import Act
>>> m = Act['memswish']()
>>> input = torch.randn(2)
>>> output = m(input)
Expand All @@ -111,10 +128,12 @@ class Mish(nn.Module):

Examples::

>>> import torch
>>> from monai.networks.layers.factories import Act
>>> m = Act['mish']()
>>> input = torch.randn(2)
>>> output = m(input)
"""

def forward(self, input: torch.Tensor) -> torch.Tensor:
return input * torch.tanh(torch.nn.functional.softplus(input))
def forward(self, input: torch.Tensor):
return monai_mish(input)