Class Sigmoid

Plot of sigmoid and its gradient

Calculates:

output = 1.0 / (1.0 + exp(-input))

Gradient:

var out = 1.0 / (1.0 + exp(-input)); inputGrad = outputGrad * out * (1.0 - out)

Hierarchy

Constructors

constructor

Accessors

[toStringTag] children immediateParameters namedChildren training

Methods

addModule buffers eval forward getBuffer loadStateDict modules namedBuffers namedModules namedParameters parameters registerBuffer requiresGrad stateDict toString train zeroGrad

Constructors

constructor

Accessors

[toStringTag]

  • get [toStringTag](): string

  • Returns string

children

  • get children(): Module[]

  • Returns the immediate children (submodules) of this module.

    Returns Module[]

Private immediateParameters

namedChildren

  • get namedChildren(): [string | number, Module][]

  • Returns the immediate children (submodules) of this module along with their names.

    Returns [string | number, Module][]

training

  • get training(): boolean

  • Returns boolean

Methods

addModule

  • Adds a child module of the current module.

    The module can be accessed as a property using the given name.

    Parameters

    • name: string | number

      name of the child module

    • module: Module

      child module to be added

    Returns void

buffers

eval

forward

getBuffer

loadStateDict

  • Loads the state of the module and its descendants from the given state dictionary.

    Parameters

    • stateDict: StateDict

      The state dictionary containing the state of the module to load.

    Returns void

modules

namedBuffers

  • Parameters

    • prefix: string = ""
    • recurse: boolean = true
    • removeDuplicate: boolean = true

    Returns Generator<[string, null | Tensor], any, unknown>

namedModules

  • Returns this module and its descendants' along with their prefixed names.

    Parameters

    • Optional memo: Set<Module>

      is a set of modules used to avoid double counting.

    • prefix: string = ""

      is prependended to the names of the modules.

    • removeDuplicate: boolean = true

      is a boolean indicating whether to remove duplicate modules.

    Returns Generator<[string | number, Module], any, unknown>

    a generator of [prefixed name, module] pairs.

namedParameters

  • Gets this module and its descendants' (if recurse = true) parameters along with their prefixed names.

    Parameters

    • prefix: string = ""

      is prependended to the names of the parameters

    • recurse: boolean = true

      whether to include submodule parameters

    • removeDuplicate: boolean = true

      whether to remove duplicate parameters

    Returns Generator<[string, Parameter], any, unknown>

    a generator of [prefixed name, parameter] pairs

parameters

registerBuffer

requiresGrad

  • Change if autograd should record operations on parameters in this module.

    Parameters

    • requiresGrad: boolean = true

      whether to enable gradient calculation for parameters in this module.

    Returns ThisType<Module>

    this module

stateDict

  • Returns the state dictionary of the module and its descendants.

    Parameters

    • Optional destination: StateDict

      An optional state dictionary to update with the module's state.

    • prefix: string = ""

      A string to prepend to the names of the state entries (default is an empty string).

    • keepVars: boolean = true

      A boolean flag, if true keeps the tensors attached to autograd (default is true).

    Returns StateDict

    The updated state dictionary containing the module's state.

toString

  • Produces a multiline readable string of the module and its descendants.

    Returns string

    A string representation of the module and its descendants.

train

  • Sets the module in training mode.

    Parameters

    • mode: boolean = true

      whether to set training mode (true) or evaluation mode (false).

    Returns ThisType<Module>

    this module

zeroGrad

  • Zeros out the gradients of all parameters. That can be accomplished either by setting the grad property to null (setToNull=true) or by filling the gradient tensor with zeros (setToNull=false).

    Parameters

    • setToNull: boolean = true

      whether to set gradients to null (true) or to zero tensors (false).

    Returns ThisType<Module>

    this module

Settings

Member Visibility

Theme

On This Page

Generated using TypeDoc