我具有以下功能
def msfe(ys, ts):
ys=ys.detach().numpy() #output from the network
ts=ts.detach().numpy() #Target (true labels)
pred_class = (ys>=0.5)
n_0 = sum(ts==0) #Number of true negatives
n_1 = sum(ts==1) #Number of true positives
FPE = sum((ts==0)[[bool(p) for p in (pred_class==1)]])/n_0 #False positive error
FNE = sum((ts==1)[[bool(p) for p in (pred_class==0)]])/n_1 #False negative error
loss= FPE**2+FNE**2
loss=torch.tensor(loss,dtype=torch.float64,requires_grad=True)
return loss
我想知道,由于ys
和ts
没有grad
标志,Pytorch中的自动分级是否正常工作。
所以我的问题是:在FPE,FNE,ys,ts,n_1,n_0
起作用之前,所有变量(optimizer.step()
)是否都必须是张量吗?或者它仅仅是最终函数(loss
)可以吗?是哪个?
答案 0 :(得分:3)
您要通过optimizer.step()
优化的所有变量都必须具有渐变。
在您的情况下,网络会y
进行预测,因此,您不应该detach
(根据图表)进行预测。
通常,您不需要更改targets
,因此这些不需要渐变。不过,您不必detach
就可以使用它们,默认情况下张量不需要梯度,并且不会向后传播。
Loss
将具有渐变(如果成分(至少一种)具有渐变)。
总体来说,您几乎不需要手动进行维护。
顺便说一句。 请勿在PyTorch中使用numpy
,这种情况很少见。您可以在PyTorch张量的numpy
数组上执行大部分操作。
BTW2。 Variable
中不再有pytorch
这样的东西,只有需要梯度的张量和不需要梯度的张量。
实际上,您正在使用不可区分的函数(即>=
和==
)。那些仅在输出时会给您带来麻烦,因为那些需要渐变(尽管您可以对==
使用>=
和targets
)。
下面我附上了损失函数,并在评论中概述了其中的问题:
# Gradient can't propagate if you detach and work in another framework
# Most Python constructs should be fine, detaching will ruin it though.
def msfe(outputs, targets):
# outputs=outputs.detach().numpy() # Do not detach, no need to do that
# targets=targets.detach().numpy() # No need for numpy either
pred_class = outputs >= 0.5 # This one is non-differentiable
# n_0 = sum(targets==0) # Do not use sum, there is pytorch function for that
# n_1 = sum(targets==1)
n_0 = torch.sum(targets == 0) # Those are not differentiable, but...
n_1 = torch.sum(targets == 1) # It does not matter as those are targets
# FPE = sum((targets==0)[[bool(p) for p in (pred_class==1)]])/n_0 # Do not use Python bools
# FNE = sum((targets==1)[[bool(p) for p in (pred_class==0)]])/n_1 # Stay within PyTorch
# Those two below are non-differentiable due to == sign as well
FPE = torch.sum((targets == 0.0) * (pred_class == 1.0)).float() / n_0
FNE = torch.sum((targets == 1.0) * (pred_class == 0.0)).float() / n_1
# This is obviously fine
loss = FPE ** 2 + FNE ** 2
# Loss should be a tensor already, don't do things like that
# Gradient will not be propagated, you will have a new tensor
# Always returning gradient of `1` and that's all
# loss = torch.tensor(loss, dtype=torch.float64, requires_grad=True)
return loss
因此,您需要除去3个不可微的部分。原则上,您可以尝试使用网络中的连续输出来对其进行近似(假设您使用sigmoid
作为激活)。这是我的看法:
def msfe_approximation(outputs, targets):
n_0 = torch.sum(targets == 0) # Gradient does not flow through it, it's okay
n_1 = torch.sum(targets == 1) # Same as above
FPE = torch.sum((targets == 0) * outputs).float() / n_0
FNE = torch.sum((targets == 1) * (1 - outputs)).float() / n_1
return FPE ** 2 + FNE ** 2
请注意,要使FPE
最小化,outputs
将尝试在zero
为零的索引上成为targets
。与FNE
类似,如果目标是1
,则网络也会尝试输出1
。
请注意,该想法与BCELoss(二进制交叉熵)相似。
最后,例如,您可以对其进行运行,仅用于完整性检查:
if __name__ == "__main__":
model = torch.nn.Sequential(
torch.nn.Linear(30, 100),
torch.nn.ReLU(),
torch.nn.Linear(100, 200),
torch.nn.ReLU(),
torch.nn.Linear(200, 1),
torch.nn.Sigmoid(),
)
optimizer = torch.optim.Adam(model.parameters())
targets = torch.randint(high=2, size=(64, 1)) # random targets
inputs = torch.rand(64, 30) # random data
for _ in range(1000):
optimizer.zero_grad()
outputs = model(inputs)
loss = msfe_approximation(outputs, targets)
print(loss)
loss.backward()
optimizer.step()
print(((model(inputs) >= 0.5) == targets).float().mean())