损失函数
def SequenceMask(X, X_len,value=0):
maxlen = X.size(1)
mask = torch.arange(maxlen)[None, :].to(X_len.device) < X_len[:, None]
X[~mask]=value
return X
X = torch.tensor([[1,2,3], [4,5,6]])
SequenceMask(X,torch.tensor([1,2]))
X = torch.ones((2,3, 4))
SequenceMask(X, torch.tensor([1,2]),value=-1)
class MaskedSoftmaxCELoss(nn.CrossEntropyLoss):
# pred shape: (batch_size, seq_len, vocab_size)
# label shape: (batch_size, seq_len)
# valid_length shape: (batch_size, )
def forward(self, pred, label, valid_length):
# the sample weights shape should be (batch_size, seq_len)
weights = torch.ones_like(label)
weights = SequenceMask(weights, valid_length).float()
self.reduction='none'
output=super(MaskedSoftmaxCELoss, self).forward(pred.transpose(1,2), label)
return (output*weights).mean(dim=1)
loss = MaskedSoftmaxCELoss()
loss(torch.ones((3, 4, 10)), torch.ones((3,4),dtype=torch.long), torch.tensor([4,3,0]))
作者：billgates2020
相关文章
Cady
2020-01-13
Tallulah
2023-07-20
Bea
2023-07-20
Ianthe
2023-07-20
Lani
2023-07-20
Thirza
2023-07-20
Rhoda
2023-07-20
Tesia
2023-07-20
Ursula
2023-07-20
Edana
2023-07-20
Dabria
2023-07-20
Paula
2023-07-20
Peony
2023-07-20
Rayna
2023-07-20
Fawn
2023-07-21
Tia
2023-07-21
Victoria
2023-07-21
Olathe
2023-07-21