Losses

Losses not available in fastai or Pytorch.

HuberLoss

 HuberLoss (reduction='mean', delta=1.0)

Huber loss

Creates a criterion that uses a squared term if the absolute element-wise error falls below delta and a delta-scaled L1 term otherwise. This loss combines advantages of both :class:L1Loss and :class:MSELoss; the delta-scaled L1 region makes the loss less sensitive to outliers than :class:MSELoss, while the L2 region provides smoothness over :class:L1Loss near 0. See Huber loss <https://en.wikipedia.org/wiki/Huber_loss>_ for more information. This loss is equivalent to nn.SmoothL1Loss when delta == 1.

source

LogCoshLoss

 LogCoshLoss (reduction='mean', delta=1.0)

Base class for all neural network modules.

Your models should also subclass this class.

Modules can also contain other Modules, allowing to nest them in a tree structure. You can assign the submodules as regular attributes::

import torch.nn as nn
import torch.nn.functional as F

class Model(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(1, 20, 5)
        self.conv2 = nn.Conv2d(20, 20, 5)

    def forward(self, x):
        x = F.relu(self.conv1(x))
        return F.relu(self.conv2(x))

Submodules assigned in this way will be registered, and will have their parameters converted too when you call :meth:to, etc.

.. note:: As per the example above, an __init__() call to the parent class must be made before assignment on the child.

:ivar training: Boolean represents whether this module is in training or evaluation mode. :vartype training: bool

inp = torch.rand(8, 3, 10)
targ = torch.randn(8, 3, 10)
test_close(HuberLoss(delta=1)(inp, targ), nn.SmoothL1Loss()(inp, targ))
LogCoshLoss()(inp, targ)

tensor(0.4588)

source

MaskedLossWrapper

 MaskedLossWrapper (crit)

Same as nn.Module, but no need for subclasses to call super().__init__

inp = torch.rand(8, 3, 10)
targ = torch.randn(8, 3, 10)
targ[targ >.8] = np.nan
nn.L1Loss()(inp, targ), MaskedLossWrapper(nn.L1Loss())(inp, targ)

(tensor(nan), tensor(1.0520))

source

CenterPlusLoss

 CenterPlusLoss (loss, c_out, λ=0.01, logits_dim=None)

Same as nn.Module, but no need for subclasses to call super().__init__

source

CenterLoss

 CenterLoss (c_out, logits_dim=None)

Code in Pytorch has been slightly modified from: https://github.com/KaiyangZhou/pytorch-center-loss/blob/master/center_loss.py Based on paper: Wen et al. A Discriminative Feature Learning Approach for Deep Face Recognition. ECCV 2016.

Args: c_out (int): number of classes. logits_dim (int): dim 1 of the logits. By default same as c_out (for one hot encoded logits)

c_in = 10
x = torch.rand(64, c_in).to(device=default_device())
x = F.softmax(x, dim=1)
label = x.max(dim=1).indices
CenterLoss(c_in).to(x.device)(x, label), CenterPlusLoss(LabelSmoothingCrossEntropyFlat(), c_in).to(x.device)(x, label)

(tensor(9.2481, grad_fn=<DivBackward0>),
 TensorBase(2.3559, grad_fn=<AliasBackward0>))

CenterPlusLoss(LabelSmoothingCrossEntropyFlat(), c_in)

CenterPlusLoss(loss=FlattenedLoss of LabelSmoothingCrossEntropy(), c_out=10, λ=0.01)

source

FocalLoss

 FocalLoss (alpha:Optional[torch.Tensor]=None, gamma:float=2.0,
            reduction:str='mean')

Weighted, multiclass focal loss

inputs = torch.normal(0, 2, (16, 2)).to(device=default_device())
targets = torch.randint(0, 2, (16,)).to(device=default_device())
FocalLoss()(inputs, targets)

tensor(0.9829)

source

TweedieLoss

 TweedieLoss (p=1.5, eps=1e-08)

Same as nn.Module, but no need for subclasses to call super().__init__

c_in = 10
output = torch.rand(64).to(device=default_device())
target = torch.rand(64).to(device=default_device())
TweedieLoss().to(output.device)(output, target)

tensor(3.0539)