NLLLoss#

classtorch.nn.NLLLoss(weight=None,size_average=None,ignore_index=-100,reduce=None,reduction='mean')[source]#

The negative log likelihood loss. It is useful to train a classificationproblem withC classes.

If provided, the optional argumentweight should be a 1D Tensor assigningweight to each of the classes. This is particularly useful when you have anunbalanced training set.

Theinput given through a forward call is expected to containlog-probabilities of each class.input has to be a Tensor of size either$(minibatch,C)$(minibatch,C) or$(minibatch,C,d_1,d_2,\mathrm{.}\mathrm{.}\mathrm{.},d_K)$(minibatch,C,d1​,d2​,...,dK​)with$K\ge 1$K≥1 for theK-dimensional case. The latter is useful forhigher dimension inputs, such as computing NLL loss per-pixel for 2D images.

Obtaining log-probabilities in a neural network is easily achieved byadding aLogSoftmax layer in the last layer of your network.You may useCrossEntropyLoss instead, if you prefer not to add an extralayer.

Thetarget that this loss expects should be a class index in the range$[0,C-1]$[0,C−1]whereC = number of classes; ifignore_index is specified, this loss also acceptsthis class index (this index may not necessarily be in the class range).

The unreduced (i.e. withreduction set to'none') loss can be described as:

$$\begin{gathered} \mathrm{\ell }(x,y)=L=\{l_1,\dots ,l_N{\}}^{\mathrm{\top }}, \\ {l}_n=-w_{y_n}{x}_{n,y_n}, \\ {w}_c=\text{weight}[c]\cdot \mathbb{1}\{c≠\text{ignore\_index}\}, \end{gathered}$$ℓ(x,y)=L={l1​,…,lN​}⊤,ln​=−wyn​​xn,yn​​,wc​=weight[c]⋅1{c=ignore_index},

where$x$x is the input,$y$y is the target,$w$w is the weight, and$N$N is the batch size. Ifreduction is not'none'(default'mean'), then

ℓ(x,y)={∑n=1N​∑n=1N​wyn​​1​ln​,∑n=1N​ln​,​if reduction=‘mean’;if reduction=‘sum’.​

Parameters

• weight (Tensor,optional) – a manual rescaling weight given to eachclass. If given, it has to be a Tensor of sizeC. Otherwise, it istreated as if having all ones.

• size_average (bool,optional) – Deprecated (seereduction). By default,the losses are averaged over each loss element in the batch. Note that forsome losses, there are multiple elements per sample. If the fieldsize_averageis set toFalse, the losses are instead summed for each minibatch. Ignoredwhenreduce isFalse. Default:None

• ignore_index (int,optional) – Specifies a target value that is ignoredand does not contribute to the input gradient. Whensize_average isTrue, the loss is averaged overnon-ignored targets.

• reduce (bool,optional) – Deprecated (seereduction). By default, thelosses are averaged or summed over observations for each minibatch dependingonsize_average. Whenreduce isFalse, returns a loss perbatch element instead and ignoressize_average. Default:None

• reduction (str,optional) – Specifies the reduction to apply to the output:'none' |'mean' |'sum'.'none': no reduction willbe applied,'mean': the weighted mean of the output is taken,'sum': the output will be summed. Note:size_averageandreduce are in the process of being deprecated, and inthe meantime, specifying either of those two args will overridereduction. Default:'mean'

Shape::

• Input:$(N,C)$(N,C) or$(C)$(C), whereC = number of classes,N = batch size, or$(N,C,d_1,d_2,\mathrm{.}\mathrm{.}\mathrm{.},d_K)$(N,C,d1​,d2​,...,dK​) with$K\ge 1$K≥1in the case ofK-dimensional loss.

• Target:$(N)$(N) or$()$(), where each value is$0\le \text{targets}[i]\le C-1$0≤targets[i]≤C−1, or$(N,d_1,d_2,\mathrm{.}\mathrm{.}\mathrm{.},d_K)$(N,d1​,d2​,...,dK​) with$K\ge 1$K≥1 in the case ofK-dimensional loss.

• Output: Ifreduction is'none', shape$(N)$(N) or$(N,d_1,d_2,\mathrm{.}\mathrm{.}\mathrm{.},d_K)$(N,d1​,d2​,...,dK​) with$K\ge 1$K≥1 in the case of K-dimensional loss.Otherwise, scalar.

Examples

>>>log_softmax=nn.LogSoftmax(dim=1)>>>loss_fn=nn.NLLLoss()>>># input to NLLLoss is of size N x C = 3 x 5>>>input=torch.randn(3,5,requires_grad=True)>>># each element in target must have 0 <= value < C>>>target=torch.tensor([1,0,4])>>>loss=loss_fn(log_softmax(input),target)>>>loss.backward()>>>>>>>>># 2D loss example (used, for example, with image inputs)>>>N,C=5,4>>>loss_fn=nn.NLLLoss()>>>data=torch.randn(N,16,10,10)>>>conv=nn.Conv2d(16,C,(3,3))>>>log_softmax=nn.LogSoftmax(dim=1)>>># output of conv forward is of shape [N, C, 8, 8]>>>output=log_softmax(conv(data))>>># each element in target must have 0 <= value < C>>>target=torch.empty(N,8,8,dtype=torch.long).random_(0,C)>>># input to NLLLoss is of size N x C x height (8) x width (8)>>>loss=loss_fn(output,target)>>>loss.backward()

forward(input,target)[source]#

Runs the forward pass.

Return type

Tensor