Breaking News: Grepper is joining You.com. Read the official announcement!

pytorch add dropout to sequential

Jason Okoro answered on July 8, 2023 Popularity 4/10 Helpfulness 2/10

answer pytorch add dropout to sequential

pytorch add dropout to sequential

Comment

class DropoutLayer(nn.Module):
    def __init__(self, p):
        super().__init__()
        self.p = p

    def forward(self, input):
        if self.training:
            u1 = (np.random.rand(*input.shape)<self.p) / self.p
            u1 *= u1
            return u1
        else:
            input *= self.p

Popularity 4/10 Helpfulness 2/10 Language python

Source: discuss.pytorch.org

Tags: add dropout python pytorch sequential

Link to this answer
Share Copy Link