gru

Documentation

Index

• type Model
  • func New[T float.DType](in, out int) *Model
  • func (m *Model) Forward(xs ...mat.Tensor) []mat.Tensor
  • func (m *Model) Next(state *State, x mat.Tensor) (s *State)
• type State

Types

type Model

type Model struct {
	nn.Module
	WPart    *nn.Param
	WPartRec *nn.Param
	BPart    *nn.Param
	WRes     *nn.Param
	WResRec  *nn.Param
	BRes     *nn.Param
	WCand    *nn.Param
	WCandRec *nn.Param
	BCand    *nn.Param
}

Model contains the serializable parameters.

func New

func New[T float.DType](in, out int) *Model

New returns a new model with parameters initialized to zeros.

func (*Model) Forward

func (m *Model) Forward(xs ...mat.Tensor) []mat.Tensor

Forward performs the forward step for each input node and returns the result.

func (*Model) Next

func (m *Model) Next(state *State, x mat.Tensor) (s *State)

Next performs a single forward step, producing a new state.

r = sigmoid(wr (dot) x + br + wrRec (dot) yPrev) p = sigmoid(wp (dot) x + bp + wpRec (dot) yPrev) c = f(wc (dot) x + bc + wcRec (dot) (yPrev * r)) y = p * c + (1 - p) * yPrev

type State

type State struct {
	R mat.Tensor
	P mat.Tensor
	C mat.Tensor
	Y mat.Tensor
}

State represent a state of the GRU recurrent network.