Python torch._C属性代码示例

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def max_unpool1d(input, indices, kernel_size, stride=None, padding=0,
                 output_size=None):
    # type: (Tensor, Tensor, BroadcastingList1[int], Optional[BroadcastingList1[int]], BroadcastingList1[int], Optional[BroadcastingList1[int]]) -> Tensor  # noqa
    r"""Computes a partial inverse of :class:`MaxPool1d`.

    See :class:`~torch.nn.MaxUnpool1d` for details.
    """
    kernel_size = _single(kernel_size)
    if stride is not None:
        _stride = _single(torch.jit._unwrap_optional(stride))
    else:
        _stride = kernel_size
    padding = _single(padding)
    output_size = _unpool_output_size(input, kernel_size, _stride, padding,
                                      output_size)
    return torch._C._nn.max_unpool2d(input.unsqueeze(3), indices.unsqueeze(3), output_size + [1]).squeeze(3) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def max_unpool2d(input, indices, kernel_size, stride=None, padding=0,
                 output_size=None):
    # type: (Tensor, Tensor, BroadcastingList2[int], Optional[BroadcastingList2[int]], BroadcastingList2[int], Optional[BroadcastingList2[int]]) -> Tensor  # noqa
    r"""Computes a partial inverse of :class:`MaxPool2d`.

    See :class:`~torch.nn.MaxUnpool2d` for details.
    """
    kernel_size = _pair(kernel_size)
    if stride is not None:
        _stride = _pair(torch.jit._unwrap_optional(stride))
    else:
        _stride = kernel_size
    padding = _pair(padding)
    output_size = _unpool_output_size(input, kernel_size, _stride, padding,
                                      output_size)
    return torch._C._nn.max_unpool2d(input, indices, output_size) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def adaptive_avg_pool3d(input, output_size):
    # type: (Tensor, BroadcastingList3[int]) -> Tensor
    r"""
    Applies a 3D adaptive average pooling over an input signal composed of
    several input planes.

    See :class:`~torch.nn.AdaptiveAvgPool3d` for details and output shape.

    Args:
        output_size: the target output size (single integer or
            triple-integer tuple)
    """
    _output_size = _list_with_default(output_size, input.size())
    return torch._C._nn.adaptive_avg_pool3d(input, _output_size)


# Activation functions 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def glu(input, dim=-1):
    # type: (Tensor, int) -> Tensor
    r"""
    glu(input, dim=-1) -> Tensor

    The gated linear unit. Computes:

    .. math ::

        H = A \times \sigma(B)

    where `input` is split in half along `dim` to form `A` and `B`.

    See `Language Modeling with Gated Convolutional Networks <https://arxiv.org/abs/1612.08083>`_.

    Args:
        input (Tensor): input tensor
        dim (int): dimension on which to split the input
    """
    if input.dim() == 0:
        raise RuntimeError("glu does not suppport scalars because halving size must be even")
    return torch._C._nn.glu(input, dim) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def smooth_l1_loss(input, target, size_average=None, reduce=None, reduction='mean'):
    # type: (Tensor, Tensor, Optional[bool], Optional[bool], str) -> Tensor
    r"""Function that uses a squared term if the absolute
    element-wise error falls below 1 and an L1 term otherwise.

    See :class:`~torch.nn.SmoothL1Loss` for details.
    """
    if size_average is not None or reduce is not None:
        reduction = _Reduction.legacy_get_string(size_average, reduce)
    if target.requires_grad:
        ret = _smooth_l1_loss(input, target)
        if reduction != 'none':
            ret = torch.mean(ret) if reduction == 'mean' else torch.sum(ret)
    else:
        expanded_input, expanded_target = torch.broadcast_tensors(input, target)
        ret = torch._C._nn.smooth_l1_loss(expanded_input, expanded_target, _Reduction.get_enum(reduction))
    return ret 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def l1_loss(input, target, size_average=None, reduce=None, reduction='mean'):
    # type: (Tensor, Tensor, Optional[bool], Optional[bool], str) -> Tensor
    r"""l1_loss(input, target, size_average=None, reduce=None, reduction='mean') -> Tensor

    Function that takes the mean element-wise absolute value difference.

    See :class:`~torch.nn.L1Loss` for details.
    """
    if size_average is not None or reduce is not None:
        reduction = _Reduction.legacy_get_string(size_average, reduce)
    if target.requires_grad:
        ret = torch.abs(input - target)
        if reduction != 'none':
            ret = torch.mean(ret) if reduction == 'mean' else torch.sum(ret)
    else:
        expanded_input, expanded_target = torch.broadcast_tensors(input, target)
        ret = torch._C._nn.l1_loss(expanded_input, expanded_target, _Reduction.get_enum(reduction))
    return ret 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def multi_margin_loss(input, target, p=1, margin=1., weight=None, size_average=None,
                      reduce=None, reduction='mean'):
    # type: (Tensor, Tensor, int, float, Optional[Tensor], Optional[bool], Optional[bool], str) -> Tensor
    r"""multi_margin_loss(input, target, p=1, margin=1, weight=None, size_average=None,
                          reduce=None, reduction='mean') -> Tensor

    See :class:`~torch.nn.MultiMarginLoss` for details.
    """
    if size_average is not None or reduce is not None:
        reduction_enum = _Reduction.legacy_get_enum(size_average, reduce)
    else:
        reduction_enum = _Reduction.get_enum(reduction)
    if p != 1 and p != 2:
        raise ValueError('only p == 1 and p == 2 supported')
    if weight is not None:
        weight = torch.jit._unwrap_optional(weight)
        if weight.dim() != 1:
            raise ValueError('weight must be one-dimensional')

    return torch._C._nn.multi_margin_loss(input, target, p, margin, weight, reduction_enum) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def max_pool2d_with_indices(input, kernel_size, stride=None, padding=0, dilation=1,
                            ceil_mode=False, return_indices=False):
    # type: (Tensor, BroadcastingList2[int], Optional[BroadcastingList2[int]], BroadcastingList2[int], BroadcastingList2[int], bool, bool) -> Tuple[Tensor, Tensor]  # noqa
    r"""Applies a 2D max pooling over an input signal composed of several input
    planes.

    See :class:`~torch.nn.MaxPool2d` for details.
    """
    if stride is None:
        _stride = torch.jit.annotate(List[int], [])
    else:
        _stride = torch.jit._unwrap_optional(stride)
    return torch._C._nn.max_pool2d_with_indices(input, kernel_size, _stride, padding, dilation, ceil_mode) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def max_pool3d_with_indices(input, kernel_size, stride=None, padding=0,
                            dilation=1, ceil_mode=False, return_indices=False):
    # type: (Tensor, BroadcastingList3[int], Optional[BroadcastingList3[int]], BroadcastingList3[int], BroadcastingList3[int], bool, bool) -> Tuple[Tensor, Tensor]  # noqa
    r"""Applies a 3D max pooling over an input signal composed of several input
    planes.

    See :class:`~torch.nn.MaxPool3d` for details.
    """
    if stride is None:
        _stride = torch.jit.annotate(List[int], [])
    else:
        _stride = torch.jit._unwrap_optional(stride)
    return torch._C._nn.max_pool3d_with_indices(
        input, kernel_size, _stride, padding, dilation, ceil_mode) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def adaptive_max_pool2d_with_indices(input, output_size, return_indices=False):
    # type: (Tensor, BroadcastingList1[int], bool) -> Tuple[Tensor, Tensor]
    r"""Applies a 2D adaptive max pooling over an input signal composed of
    several input planes.

    See :class:`~torch.nn.AdaptiveMaxPool2d` for details and output shape.

    Args:
        output_size: the target output size (single integer or
            double-integer tuple)
        return_indices: whether to return pooling indices. Default: ``False``
    """
    output_size = _list_with_default(output_size, input.size())
    return torch._C._nn.adaptive_max_pool2d(input, output_size) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def adaptive_max_pool3d_with_indices(input, output_size, return_indices=False):
    # type: (Tensor, BroadcastingList1[int], bool) -> Tuple[Tensor, Tensor]
    r"""Applies a 3D adaptive max pooling over an input signal composed of
    several input planes.

    See :class:`~torch.nn.AdaptiveMaxPool3d` for details and output shape.

    Args:
        output_size: the target output size (single integer or
            triple-integer tuple)
        return_indices: whether to return pooling indices. Default: ``False``
    """
    output_size = _list_with_default(output_size, input.size())
    return torch._C._nn.adaptive_max_pool3d(input, output_size) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def adaptive_avg_pool2d(input, output_size):
    # type: (Tensor, BroadcastingList2[int]) -> Tensor
    r"""
    Applies a 2D adaptive average pooling over an input signal composed of
    several input planes.

    See :class:`~torch.nn.AdaptiveAvgPool2d` for details and output shape.

    Args:
        output_size: the target output size (single integer or
            double-integer tuple)
    """
    _output_size = _list_with_default(output_size, input.size())
    return torch._C._nn.adaptive_avg_pool2d(input, _output_size) 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def elu(input, alpha=1., inplace=False):
    # type: (Tensor, float, bool) -> Tensor
    r"""Applies element-wise,
    :math:`\text{ELU}(x) = \max(0,x) + \min(0, \alpha * (\exp(x) - 1))`.

    See :class:`~torch.nn.ELU` for more details.
    """
    if inplace:
        result = torch._C._nn.elu_(input, alpha)
    else:
        result = torch._C._nn.elu(input, alpha)
    return result 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def leaky_relu(input, negative_slope=0.01, inplace=False):
    # type: (Tensor, float, bool) -> Tensor
    r"""
    leaky_relu(input, negative_slope=0.01, inplace=False) -> Tensor

    Applies element-wise,
    :math:`\text{LeakyReLU}(x) = \max(0, x) + \text{negative\_slope} * \min(0, x)`

    See :class:`~torch.nn.LeakyReLU` for more details.
    """
    if inplace:
        result = torch._C._nn.leaky_relu_(input, negative_slope)
    else:
        result = torch._C._nn.leaky_relu(input, negative_slope)
    return result 

# 需要导入模块: import torch [as 别名]
# 或者: from torch import _C [as 别名]
def multilabel_margin_loss(input, target, size_average=None, reduce=None, reduction='mean'):
    # type: (Tensor, Tensor, Optional[bool], Optional[bool], str) -> Tensor
    r"""multilabel_margin_loss(input, target, size_average=None, reduce=None, reduction='mean') -> Tensor

    See :class:`~torch.nn.MultiLabelMarginLoss` for details.
    """
    if size_average is not None or reduce is not None:
        reduction_enum = _Reduction.legacy_get_enum(size_average, reduce)
    else:
        reduction_enum = _Reduction.get_enum(reduction)
    return torch._C._nn.multilabel_margin_loss(input, target, reduction_enum)