Documentation
¶
Overview ¶
Package loss provides various loss functions for neural networks.
Index ¶
- type CrossEntropyLoss
- func (cel *CrossEntropyLoss[T]) Backward(ctx context.Context, dOut *tensor.TensorNumeric[T]) ([]*tensor.TensorNumeric[T], error)
- func (cel *CrossEntropyLoss[T]) Forward(ctx context.Context, predictions *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)
- func (cel *CrossEntropyLoss[T]) OutputShape() []int
- func (cel *CrossEntropyLoss[T]) Parameters() []graph.Parameter[T]
- type Loss
- type MSE
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type CrossEntropyLoss ¶
CrossEntropyLoss computes the cross-entropy loss.
func NewCrossEntropyLoss ¶
func NewCrossEntropyLoss[T tensor.Numeric](engine compute.Engine[T]) *CrossEntropyLoss[T]
NewCrossEntropyLoss creates a new CrossEntropyLoss layer.
func (*CrossEntropyLoss[T]) Backward ¶
func (cel *CrossEntropyLoss[T]) Backward(ctx context.Context, dOut *tensor.TensorNumeric[T]) ([]*tensor.TensorNumeric[T], error)
Backward computes the gradients for CrossEntropyLoss. dOut is typically a scalar (1.0) for loss functions.
func (*CrossEntropyLoss[T]) Forward ¶
func (cel *CrossEntropyLoss[T]) Forward(ctx context.Context, predictions *tensor.TensorNumeric[T], targets *tensor.TensorNumeric[int]) (*tensor.TensorNumeric[T], error)
Forward computes the cross-entropy loss. Inputs: predictions (logits), targets (int labels).
func (*CrossEntropyLoss[T]) OutputShape ¶
func (cel *CrossEntropyLoss[T]) OutputShape() []int
OutputShape returns the output shape of the loss (a scalar).
func (*CrossEntropyLoss[T]) Parameters ¶
func (cel *CrossEntropyLoss[T]) Parameters() []graph.Parameter[T]
Parameters returns an empty slice as CrossEntropyLoss has no trainable parameters.
type Loss ¶
type Loss[T tensor.Numeric] interface { // Forward computes the loss and its gradient. Forward(ctx context.Context, predictions, targets *tensor.TensorNumeric[T]) (T, *tensor.TensorNumeric[T], error) }
Loss defines the interface for loss functions.
type MSE ¶
MSE calculates the mean squared error between predictions and targets.
func (*MSE[T]) Backward ¶
func (m *MSE[T]) Backward(ctx context.Context, predictions, targets *tensor.TensorNumeric[T]) (*tensor.TensorNumeric[T], error)
Backward computes the initial gradient of the loss.
func (*MSE[T]) Forward ¶
func (m *MSE[T]) Forward(ctx context.Context, predictions, targets *tensor.TensorNumeric[T]) (*tensor.TensorNumeric[T], error)
Forward computes the loss value.
Source Files