AvgPool1d#

classtorch.nn.AvgPool1d(kernel_size,stride=None,padding=0,ceil_mode=False,count_include_pad=True)[source]#

Applies a 1D average pooling over an input signal composed of several input planes.

In the simplest case, the output value of the layer with input size$(N,C,L)$(N,C,L),output$(N,C,L_{out})$(N,C,Lout​) andkernel_size$k$kcan be precisely described as:

$$\text{out}(N_i,C_j,l)=\frac{1}{k}\sum _{m=0}^{k-1}\text{input}(N_i,C_j,\text{stride}\times l+m)$$out(Ni​,Cj​,l)=k1​m=0∑k−1​input(Ni​,Cj​,stride×l+m)

Ifpadding is non-zero, then the input is implicitly zero-padded on both sidesforpadding number of points.

Note

When ceil_mode=True, sliding windows are allowed to go off-bounds if they start within the left paddingor the input. Sliding windows that would start in the right padded region are ignored.

Note

pad should be at most half of effective kernel size.

The parameterskernel_size,stride,padding can each beanint or a one-element tuple.

Parameters

• kernel_size (Union[int,tuple[int]]) – the size of the window

• stride (Union[int,tuple[int]]) – the stride of the window. Default value iskernel_size

• padding (Union[int,tuple[int]]) – implicit zero padding to be added on both sides

• ceil_mode (bool) – when True, will useceil instead offloor to compute the output shape

• count_include_pad (bool) – when True, will include the zero-padding in the averaging calculation

Shape:

• Input:$(N,C,L_{in})$(N,C,Lin​) or$(C,L_{in})$(C,Lin​).

• Output:$(N,C,L_{out})$(N,C,Lout​) or$(C,L_{out})$(C,Lout​), where

  $$L_{out}=\lfloor \frac{L_{in}+2\times \text{padding}-\text{kernel\_size}}{\text{stride}}+1\rfloor$$Lout​=⌊strideLin​+2×padding−kernel_size​+1⌋

  Per the note above, ifceil_mode is True and$(L_{out}-1)\times \text{stride}\ge L_{in}+\text{padding}$(Lout​−1)×stride≥Lin​+padding, we skip the last window as it would start in the right padded region, resulting in$L_{out}$Lout​ being reduced by one.

Examples:

>>># pool with window of size=3, stride=2>>>m=nn.AvgPool1d(3,stride=2)>>>m(torch.tensor([[[1.,2,3,4,5,6,7]]]))tensor([[[2., 4., 6.]]])

forward(input)[source]#

Runs the forward pass.

Return type

Tensor