DiffSharp

Model Type

Namespace: DiffSharp.Model

Assembly: DiffSharp.Core.dll


Base Type: obj

Record fields

Record Field Description
mode
Full Usage:
mode
Signature: Mode
Modifiers: mutable

Constructors

Constructor Description
Model()
Full Usage:
Model()
Return Type: Model
Signature: unit -> Model
Return Type: Model

Instance members

Instance member Description
this.ParametersDict
Full Usage:
this.ParametersDict
Return Type: ParameterDict
Signature: ParameterDict
Return Type: ParameterDict
this.SubModelsDict
Full Usage:
this.SubModelsDict
Return Type: Dictionary<string, Model>
Signature: Dictionary<string, Model>
Return Type: Dictionary<string, Model>
this.add(parameters, ?names)
Full Usage:
this.add(parameters, ?names) Parameter Types:
  • parameters : seq<obj>
  • ?names : seq<string>

Signature: (parameters: seq<obj> * names: seq<string>) -> unit
Parameter Types:
  • parameters : seq<obj>
  • ?names : seq<string>
this.allModels
Full Usage:
this.allModels
Return Type: Model list
Signature: Model list
Return Type: Model list
this.clone()
Full Usage:
this.clone()
Return Type: Model
Signature: unit -> Model
Return Type: Model
this.eval()
Full Usage:
this.eval()
Signature: unit -> unit
this.forward arg1
Full Usage:
this.forward arg1 Parameter Types:
  • arg0 : Tensor

Return Type: Tensor
Signature: Tensor -> Tensor
Modifiers: abstract
Parameter Types:
  • arg0 : Tensor
Return Type: Tensor
this.forwardCompose f input parameters
Full Usage:
this.forwardCompose f input parameters Parameter Types:
  • f : Tensor -> Tensor
  • input : Tensor
  • parameters : Tensor

Return Type: Tensor
Signature: (f: (Tensor -> Tensor)) -> (input: Tensor) -> (parameters: Tensor) -> Tensor
Parameter Types:
  • f : Tensor -> Tensor
  • input : Tensor
  • parameters : Tensor
Return Type: Tensor
this.forwardDiff derivatives
Full Usage:
this.forwardDiff derivatives Parameter Types:
  • derivatives : ParameterDict

Signature: (derivatives: ParameterDict) -> unit
Parameter Types:
  • derivatives : ParameterDict
this.forwardLoss f input target parameters
Full Usage:
this.forwardLoss f input target parameters Parameter Types:
  • f : Tensor -> Tensor -> Tensor
  • input : Tensor
  • target : Tensor
  • parameters : Tensor

Return Type: Tensor
Signature: (f: (Tensor -> Tensor -> Tensor)) -> (input: Tensor) -> (target: Tensor) -> (parameters: Tensor) -> Tensor
Parameter Types:
  • f : Tensor -> Tensor -> Tensor
  • input : Tensor
  • target : Tensor
  • parameters : Tensor
Return Type: Tensor
this.forwardParameters input parameters
Full Usage:
this.forwardParameters input parameters Parameter Types:
  • input : Tensor
  • parameters : Tensor

Return Type: Tensor
Signature: (input: Tensor) -> (parameters: Tensor) -> Tensor
Parameter Types:
  • input : Tensor
  • parameters : Tensor
Return Type: Tensor
this.loadParameters fileName
Full Usage:
this.loadParameters fileName Parameter Types:
  • fileName : string

Signature: (fileName: string) -> unit
Parameter Types:
  • fileName : string
this.move(?dtype, ?device, ?backend)
Full Usage:
this.move(?dtype, ?device, ?backend) Parameter Types:
  • ?dtype : Dtype
  • ?device : Device
  • ?backend : Backend

Signature: (dtype: Dtype * device: Device * backend: Backend) -> unit
Parameter Types:
  • ?dtype : Dtype
  • ?device : Device
  • ?backend : Backend
this.noDiff()
Full Usage:
this.noDiff()
Signature: unit -> unit
this.nparameters
Full Usage:
this.nparameters
Return Type: int
Signature: int
Return Type: int
this.parameters()
Full Usage:
this.parameters()
Return Type: Tensor
Signature: unit -> Tensor
Return Type: Tensor
this.parameters()
Full Usage:
this.parameters()
Signature: unit -> unit
this.parametersDict()
Full Usage:
this.parametersDict()
Return Type: ParameterDict
Signature: unit -> ParameterDict
Return Type: ParameterDict
this.parametersDict()
Full Usage:
this.parametersDict()
Signature: unit -> unit
this.reverseDiff()
Full Usage:
this.reverseDiff()
Signature: unit -> unit
this.save fileName
Full Usage:
this.save fileName Parameter Types:
  • fileName : string

Signature: (fileName: string) -> unit
Parameter Types:
  • fileName : string
this.saveParameters fileName
Full Usage:
this.saveParameters fileName Parameter Types:
  • fileName : string

Signature: (fileName: string) -> unit
Parameter Types:
  • fileName : string
this.train()
Full Usage:
this.train()
Signature: unit -> unit

Static members

Static member Description
Model.( --> )(t, m)
Full Usage:
Model.( --> )(t, m) Parameter Types:
  • t : Tensor
  • m : Model

Return Type: Tensor
Signature: (t: Tensor * m: Model) -> Tensor
Parameter Types:
  • t : Tensor
  • m : Model
Return Type: Tensor
Model.( --> )(f, m)
Full Usage:
Model.( --> )(f, m) Parameter Types:
  • f : Tensor -> Tensor
  • m : Model

Return Type: Model
Signature: (f: (Tensor -> Tensor) * m: Model) -> Model
Parameter Types:
  • f : Tensor -> Tensor
  • m : Model
Return Type: Model
Model.( --> )(m, f)
Full Usage:
Model.( --> )(m, f) Parameter Types:
  • m : Model
  • f : Tensor -> Tensor

Return Type: Model
Signature: (m: Model * f: (Tensor -> Tensor)) -> Model
Parameter Types:
  • m : Model
  • f : Tensor -> Tensor
Return Type: Model
Model.( --> )(m1, m2)
Full Usage:
Model.( --> )(m1, m2) Parameter Types:
  • m1 : Model
  • m2 : Model

Return Type: Model
Signature: (m1: Model * m2: Model) -> Model
Parameter Types:
  • m1 : Model
  • m2 : Model
Return Type: Model
Model.compose m1 m2
Full Usage:
Model.compose m1 m2 Parameter Types:
  • m1 : Model
  • m2 : Model

Return Type: Model
Signature: (m1: Model) -> (m2: Model) -> Model
Parameter Types:
  • m1 : Model
  • m2 : Model
Return Type: Model
Model.create ps f
Full Usage:
Model.create ps f Parameter Types:
  • ps : seq<obj>
  • f : Tensor -> Tensor

Return Type: Model
Signature: (ps: seq<obj>) -> (f: (Tensor -> Tensor)) -> Model
Parameter Types:
  • ps : seq<obj>
  • f : Tensor -> Tensor
Return Type: Model
Model.load fileName
Full Usage:
Model.load fileName Parameter Types:
  • fileName : string

Return Type: Model
Signature: (fileName: string) -> Model
Parameter Types:
  • fileName : string
Return Type: Model