API Reference
This page lists the public API of MathOptAI.
This page is an unstructured list of the MathOptAI API. For a more structured overview, read the Manual or Tutorial parts of this documentation.
Load all of the public the API into the current scope with:
using MathOptAIAlternatively, load only the module with:
import MathOptAIand then prefix all calls with MathOptAI. to create MathOptAI.<NAME>.
AbstractPredictor
MathOptAI.AbstractPredictor — Typeabstract type AbstractPredictor endAn abstract type representing different types of prediction models.
Methods
All subtypes must implement:
add_predictor
MathOptAI.add_predictor — Functionadd_predictor(
+API Reference · MathOptAI.jl API Reference
This page lists the public API of MathOptAI.
Info This page is an unstructured list of the MathOptAI API. For a more structured overview, read the Manual or Tutorial parts of this documentation.
Load all of the public the API into the current scope with:
using MathOptAI
Alternatively, load only the module with:
import MathOptAI
and then prefix all calls with MathOptAI. to create MathOptAI.<NAME>.
AbstractPredictor
MathOptAI.AbstractPredictor — Typeabstract type AbstractPredictor end
An abstract type representing different types of prediction models.
Methods
All subtypes must implement:
sourceadd_predictor
MathOptAI.add_predictor — Functionadd_predictor(
model::JuMP.AbstractModel,
predictor::AbstractPredictor,
x::Vector,
@@ -23,7 +23,7 @@
├ variables [1]
│ └ moai_Affine[1]
└ constraints [1]
- └ 2 x[1] + 3 x[2] - moai_Affine[1] = 0
sourcebuild_predictor
MathOptAI.build_predictor — Methodbuild_predictor(extension; kwargs...)::AbstractPredictor
A uniform interface to convert various extension types to an AbstractPredictor.
See the various extension docstrings for details.
sourceAffine
MathOptAI.Affine — TypeAffine(
+ └ 2 x[1] + 3 x[2] - moai_Affine[1] = 0
sourcebuild_predictor
MathOptAI.build_predictor — Methodbuild_predictor(extension; kwargs...)::AbstractPredictor
A uniform interface to convert various extension types to an AbstractPredictor.
See the various extension docstrings for details.
sourceAffine
MathOptAI.Affine — TypeAffine(
A::Matrix{T},
b::Vector{T} = zeros(T, size(A, 1)),
) where {T} <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = A x + b\]
Example
julia> using JuMP, MathOptAI
@@ -60,7 +60,7 @@
julia> formulation
ReducedSpace(Affine(A, b) [input: 2, output: 1])
├ variables [0]
-└ constraints [0]
sourceBinaryDecisionTree
MathOptAI.BinaryDecisionTree — TypeBinaryDecisionTree{K,V}(
+└ constraints [0]
sourceBinaryDecisionTree
MathOptAI.BinaryDecisionTree — TypeBinaryDecisionTree{K,V}(
feat_id::Int,
feat_value::K,
lhs::Union{V,BinaryDecisionTree{K,V}},
@@ -99,7 +99,7 @@
├ moai_BinaryDecisionTree_z[2] --> {x[1] ≤ 1}
├ moai_BinaryDecisionTree_z[3] --> {x[1] ≥ 0}
├ moai_BinaryDecisionTree_z[3] --> {x[1] ≥ 1}
- └ moai_BinaryDecisionTree_z[1] - moai_BinaryDecisionTree_z[3] + moai_BinaryDecisionTree_value = 0
sourceGrayBox
MathOptAI.GrayBox — TypeGrayBox(
+ └ moai_BinaryDecisionTree_z[1] - moai_BinaryDecisionTree_z[3] + moai_BinaryDecisionTree_value = 0
sourceGrayBox
MathOptAI.GrayBox — TypeGrayBox(
output_size::Function,
callback::Function;
has_hessian::Bool = false,
@@ -141,7 +141,7 @@
julia> formulation
ReducedSpace(GrayBox)
├ variables [0]
-└ constraints [0]
sourcePipeline
MathOptAI.Pipeline — TypePipeline(layers::Vector{AbstractPredictor}) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = (l_1 \circ \ldots \circ l_N)(x)\]
where $l_i$ are a list of other AbstractPredictors.
Example
julia> using JuMP, MathOptAI
+└ constraints [0]
sourcePipeline
MathOptAI.Pipeline — TypePipeline(layers::Vector{AbstractPredictor}) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = (l_1 \circ \ldots \circ l_N)(x)\]
where $l_i$ are a list of other AbstractPredictors.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -175,11 +175,11 @@
├ moai_ReLU[1] ≥ 0
├ moai_z[1] ≥ 0
├ moai_Affine[1] - moai_ReLU[1] + moai_z[1] = 0
- └ moai_ReLU[1]*moai_z[1] = 0
sourcePytorchModel
MathOptAI.PytorchModel — TypePytorchModel(filename::String)
A wrapper struct for loading a PyTorch model.
The only supported file extension is .pt, where the .pt file has been created using torch.save(model, filename).
Warning To use PytorchModel, your code must load the PythonCall package:
import PythonCall
Example
julia> using MathOptAI
+ └ moai_ReLU[1]*moai_z[1] = 0
sourcePytorchModel
MathOptAI.PytorchModel — TypePytorchModel(filename::String)
A wrapper struct for loading a PyTorch model.
The only supported file extension is .pt, where the .pt file has been created using torch.save(model, filename).
Warning To use PytorchModel, your code must load the PythonCall package:
import PythonCall
Example
julia> using MathOptAI
julia> using PythonCall # This line is important!
-julia> predictor = PytorchModel("model.pt");
sourceQuantile
MathOptAI.Quantile — TypeQuantile{D}(distribution::D, quantiles::Vector{Float64}) where {D}
An AbstractPredictor that represents the quantiles of distribution.
Example
julia> using JuMP, Distributions, MathOptAI
+julia> predictor = PytorchModel("model.pt");
sourceQuantile
MathOptAI.Quantile — TypeQuantile{D}(distribution::D, quantiles::Vector{Float64}) where {D}
An AbstractPredictor that represents the quantiles of distribution.
Example
julia> using JuMP, Distributions, MathOptAI
julia> model = Model();
@@ -204,7 +204,7 @@
│ └ moai_quantile[2]
└ constraints [2]
├ moai_quantile[1] - op_quantile_0.1(x) = 0
- └ moai_quantile[2] - op_quantile_0.9(x) = 0
sourceReducedSpace
MathOptAI.ReducedSpace — TypeReducedSpace(predictor::AbstractPredictor)
A wrapper type for other predictors that implement a reduced-space formulation.
Example
julia> using JuMP, MathOptAI
+ └ moai_quantile[2] - op_quantile_0.9(x) = 0
sourceReducedSpace
MathOptAI.ReducedSpace — TypeReducedSpace(predictor::AbstractPredictor)
A wrapper type for other predictors that implement a reduced-space formulation.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -217,7 +217,7 @@
julia> y
2-element Vector{NonlinearExpr}:
max(0.0, x[1])
- max(0.0, x[2])
sourceReLU
MathOptAI.ReLU — TypeReLU() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
as a non-smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
+ max(0.0, x[2])
sourceReLU
MathOptAI.ReLU — TypeReLU() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
as a non-smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -257,7 +257,7 @@
julia> formulation
ReducedSpace(ReLU())
├ variables [0]
-└ constraints [0]
sourceReLUBigM
MathOptAI.ReLUBigM — TypeReLUBigM(M::Float64) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
via the big-M MIP reformulation:
\[\begin{aligned}
+└ constraints [0]
sourceReLUBigM
MathOptAI.ReLUBigM — TypeReLUBigM(M::Float64) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
via the big-M MIP reformulation:
\[\begin{aligned}
y \ge 0 \\
y \ge x \\
y \le M z \\
@@ -298,7 +298,7 @@
├ moai_z[2] binary
├ -x[2] + moai_ReLU[2] ≥ 0
├ moai_ReLU[2] - 2 moai_z[2] ≤ 0
- └ -x[2] + moai_ReLU[2] + 3 moai_z[2] ≤ 3
sourceReLUQuadratic
MathOptAI.ReLUQuadratic — TypeReLUQuadratic() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
by the reformulation:
\[\begin{aligned}
+ └ -x[2] + moai_ReLU[2] + 3 moai_z[2] ≤ 3
sourceReLUQuadratic
MathOptAI.ReLUQuadratic — TypeReLUQuadratic() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
by the reformulation:
\[\begin{aligned}
x = y - z \\
y \cdot z = 0 \\
y, z \ge 0
@@ -337,7 +337,7 @@
├ x[1] - moai_ReLU[1] + moai_z[1] = 0
├ x[2] - moai_ReLU[2] + moai_z[2] = 0
├ moai_ReLU[1]*moai_z[1] = 0
- └ moai_ReLU[2]*moai_z[2] = 0
sourceReLUSOS1
MathOptAI.ReLUSOS1 — TypeReLUSOS1() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
by the reformulation:
\[\begin{aligned}
+ └ moai_ReLU[2]*moai_z[2] = 0
sourceReLUSOS1
MathOptAI.ReLUSOS1 — TypeReLUSOS1() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \max\{0, x\}\]
by the reformulation:
\[\begin{aligned}
x = y - z \\
[y, z] \in SOS1 \\
y, z \ge 0
@@ -374,7 +374,7 @@
├ x[1] - moai_ReLU[1] + moai_z[1] = 0
├ x[2] - moai_ReLU[2] + moai_z[2] = 0
├ [moai_ReLU[1], moai_z[1]] ∈ MathOptInterface.SOS1{Float64}([1.0, 2.0])
- └ [moai_ReLU[2], moai_z[2]] ∈ MathOptInterface.SOS1{Float64}([1.0, 2.0])
sourceScale
MathOptAI.Scale — TypeScale(
+ └ [moai_ReLU[2], moai_z[2]] ∈ MathOptInterface.SOS1{Float64}([1.0, 2.0])
sourceScale
MathOptAI.Scale — TypeScale(
scale::Vector{T},
bias::Vector{T},
) where {T} <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = Diag(scale)x + bias\]
Example
julia> using JuMP, MathOptAI
@@ -417,7 +417,7 @@
julia> formulation
ReducedSpace(Scale(scale, bias))
├ variables [0]
-└ constraints [0]
sourceSigmoid
MathOptAI.Sigmoid — TypeSigmoid() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{1}{1 + e^{-x}}\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
+└ constraints [0]
sourceSigmoid
MathOptAI.Sigmoid — TypeSigmoid() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{1}{1 + e^{-x}}\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -457,7 +457,7 @@
julia> formulation
ReducedSpace(Sigmoid())
├ variables [0]
-└ constraints [0]
sourceSoftMax
MathOptAI.SoftMax — TypeSoftMax() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{e^{x}}{||e^{x}||_1}\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
+└ constraints [0]
sourceSoftMax
MathOptAI.SoftMax — TypeSoftMax() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{e^{x}}{||e^{x}||_1}\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -503,7 +503,7 @@
│ └ moai_SoftMax_denom
└ constraints [2]
├ moai_SoftMax_denom ≥ 0
- └ moai_SoftMax_denom - (0.0 + exp(x[2]) + exp(x[1])) = 0
sourceSoftPlus
MathOptAI.SoftPlus — TypeSoftPlus(; beta = 1.0) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{1}{\beta} \log(1 + e^{\beta x})\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
+ └ moai_SoftMax_denom - (0.0 + exp(x[2]) + exp(x[1])) = 0
sourceSoftPlus
MathOptAI.SoftPlus — TypeSoftPlus(; beta = 1.0) <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \frac{1}{\beta} \log(1 + e^{\beta x})\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -543,7 +543,7 @@
julia> formulation
ReducedSpace(SoftPlus(2.0))
├ variables [0]
-└ constraints [0]
sourceTanh
MathOptAI.Tanh — TypeTanh() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \tanh(x)\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
+└ constraints [0]
sourceTanh
MathOptAI.Tanh — TypeTanh() <: AbstractPredictor
An AbstractPredictor that represents the relationship:
\[y = \tanh(x)\]
as a smooth nonlinear constraint.
Example
julia> using JuMP, MathOptAI
julia> model = Model();
@@ -583,14 +583,14 @@
julia> formulation
ReducedSpace(Tanh())
├ variables [0]
-└ constraints [0]
sourceAbstractFormulation
MathOptAI.AbstractFormulation — Typeabstract type AbstractFormulation end
An abstract type representing different formulations.
sourceFormulation
MathOptAI.Formulation — Typestruct Formulation{P<:AbstractPredictor} <: AbstractFormulation
+└ constraints [0]
sourceAbstractFormulation
MathOptAI.AbstractFormulation — Typeabstract type AbstractFormulation end
An abstract type representing different formulations.
sourceFormulation
MathOptAI.Formulation — Typestruct Formulation{P<:AbstractPredictor} <: AbstractFormulation
predictor::P
variables::Vector{Any}
constraints::Vector{Any}
-end
Fields
predictor: the predictor object used to build the formulationvariables: a vector of new decision variables added to the modelconstraints: a vector of new constraints added to the model
Check the docstring of the predictor for an explanation of the formulation and the order of the elements in .variables and .constraints.
sourcePipelineFormulation
MathOptAI.PipelineFormulation — Typestruct PipelineFormulation{P<:AbstractPredictor} <: AbstractFormulation
+end
Fields
predictor: the predictor object used to build the formulationvariables: a vector of new decision variables added to the modelconstraints: a vector of new constraints added to the model
Check the docstring of the predictor for an explanation of the formulation and the order of the elements in .variables and .constraints.
sourcePipelineFormulation
MathOptAI.PipelineFormulation — Typestruct PipelineFormulation{P<:AbstractPredictor} <: AbstractFormulation
predictor::P
layers::Vector{Any}
-end
Fields
predictor: the predictor object used to build the formulationlayers: the formulation associated with each of the layers in the pipeline
sourceExtensions
MathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+end
Fields
predictor: the predictor object used to build the formulationlayers: the formulation associated with each of the layers in the pipeline
sourceExtensions
MathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::MathOptAI.Quantile{<:AbstractGPs.PosteriorGP},
x::Vector,
@@ -618,7 +618,7 @@
moai_quantile[2]
julia> @objective(model, Max, y[2] - y[1])
-moai_quantile[2] - moai_quantile[1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+moai_quantile[2] - moai_quantile[1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::Union{DecisionTree.Root,DecisionTree.DecisionTreeClassifier},
x::Vector,
@@ -647,7 +647,7 @@
julia> y
1-element Vector{VariableRef}:
- moai_BinaryDecisionTree_value
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(predictor::DecisionTree.Root)
Convert a binary decision tree from DecisionTree.jl to a BinaryDecisionTree.
Example
julia> using MathOptAI, DecisionTree
+ moai_BinaryDecisionTree_value
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(predictor::DecisionTree.Root)
Convert a binary decision tree from DecisionTree.jl to a BinaryDecisionTree.
Example
julia> using MathOptAI, DecisionTree
julia> truth(x::Vector) = x[1] <= 0.5 ? -2 : (x[2] <= 0.3 ? 3 : 4)
truth (generic function with 1 method)
@@ -665,7 +665,7 @@
Depth: 2
julia> predictor = MathOptAI.build_predictor(tree)
-BinaryDecisionTree{Float64,Int64} [leaves=3, depth=2]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+BinaryDecisionTree{Float64,Int64} [leaves=3, depth=2]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::Flux.Chain,
x::Vector;
@@ -689,7 +689,7 @@
julia> y
1-element Vector{VariableRef}:
- moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
+ moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
predictor::Flux.Chain;
config::Dict = Dict{Any,Any}(),
gray_box::Bool = false,
@@ -714,7 +714,7 @@
Pipeline with layers:
* Affine(A, b) [input: 1, output: 16]
* ReLUQuadratic()
- * Affine(A, b) [input: 16, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+ * Affine(A, b) [input: 16, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::GLM.GeneralizedLinearModel{
GLM.GlmResp{Vector{Float64},GLM.Bernoulli{Float64},GLM.LogitLink},
@@ -741,7 +741,7 @@
julia> y
1-element Vector{VariableRef}:
- moai_Sigmoid[1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+ moai_Sigmoid[1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::GLM.LinearModel,
x::Vector;
@@ -760,7 +760,7 @@
julia> y
1-element Vector{VariableRef}:
- moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
+ moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
predictor::GLM.GeneralizedLinearModel{
GLM.GlmResp{Vector{Float64},GLM.Bernoulli{Float64},GLM.LogitLink},
};
@@ -774,14 +774,14 @@
julia> predictor = MathOptAI.build_predictor(model)
Pipeline with layers:
* Affine(A, b) [input: 2, output: 1]
- * Sigmoid()
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(predictor::GLM.LinearModel)
Convert a trained linear model from GLM.jl to an Affine layer.
Example
julia> using GLM, MathOptAI
+ * Sigmoid()
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(predictor::GLM.LinearModel)
Convert a trained linear model from GLM.jl to an Affine layer.
Example
julia> using GLM, MathOptAI
julia> X, Y = rand(10, 2), rand(10);
julia> model = GLM.lm(X, Y);
julia> predictor = MathOptAI.build_predictor(model)
-Affine(A, b) [input: 2, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+Affine(A, b) [input: 2, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::Tuple{<:Lux.Chain,<:NamedTuple,<:NamedTuple},
x::Vector;
@@ -815,7 +815,7 @@
julia> y
1-element Vector{VariableRef}:
- moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
+ moai_Affine[1]
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
predictor::Tuple{<:Lux.Chain,<:NamedTuple,<:NamedTuple};
config::Dict = Dict{Any,Any}(),
)
Convert a trained neural network from Lux.jl to a Pipeline.
Supported layers
Lux.DenseLux.Scale
Supported activation functions
Lux.reluLux.sigmoidLux.softplusLux.softmaxLux.tanh
Keyword arguments
config: a dictionary that maps supported Lux activation functions to AbstractPredictors that control how the activation functions are reformulated. For example, Lux.sigmoid => MathOptAI.Sigmoid() or Lux.relu => MathOptAI.QuadraticReLU().
Example
julia> using Lux, MathOptAI, Random
@@ -847,7 +847,7 @@
Pipeline with layers:
* Affine(A, b) [input: 1, output: 16]
* ReLUQuadratic()
- * Affine(A, b) [input: 16, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+ * Affine(A, b) [input: 16, output: 1]
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::MathOptAI.PytorchModel,
x::Vector;
@@ -856,13 +856,13 @@
gray_box::Bool = false,
gray_box_hessian::Bool = false,
gray_box_device::String = "cpu",
-)
Add a trained neural network from PyTorch via PythonCall.jl to model.
Supported layers
nn.Linearnn.ReLUnn.Sequentialnn.Sigmoidnn.Softplusnn.Tanh
Keyword arguments
config: a dictionary that maps Symbols to AbstractPredictors that control how the activation functions are reformulated. For example, :Sigmoid => MathOptAI.Sigmoid() or :ReLU => MathOptAI.QuadraticReLU(). The supported Symbols are :ReLU, :Sigmoid, :SoftPlus, and :Tanh.gray_box: if true, the neural network is added as a user-defined nonlinear operator, with gradients provided by torch.func.jacrev.gray_box_hessian: if true, the gray box additionally computes the Hessian of the output using torch.func.hessian.gray_box_device: device used to construct PyTorch tensors, e.g. "cuda" to run on an Nvidia GPU.
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
+)
Add a trained neural network from PyTorch via PythonCall.jl to model.
Supported layers
nn.Linearnn.ReLUnn.Sequentialnn.Sigmoidnn.Softplusnn.Tanh
Keyword arguments
config: a dictionary that maps Symbols to AbstractPredictors that control how the activation functions are reformulated. For example, :Sigmoid => MathOptAI.Sigmoid() or :ReLU => MathOptAI.QuadraticReLU(). The supported Symbols are :ReLU, :Sigmoid, :SoftPlus, and :Tanh.gray_box: if true, the neural network is added as a user-defined nonlinear operator, with gradients provided by torch.func.jacrev.gray_box_hessian: if true, the gray box additionally computes the Hessian of the output using torch.func.hessian.gray_box_device: device used to construct PyTorch tensors, e.g. "cuda" to run on an Nvidia GPU.
sourceMathOptAI.build_predictor — MethodMathOptAI.build_predictor(
predictor::MathOptAI.PytorchModel;
config::Dict = Dict{Any,Any}(),
gray_box::Bool = false,
gray_box_hessian::Bool = false,
gray_box_device::String = "cpu",
-)
Convert a trained neural network from PyTorch via PythonCall.jl to a Pipeline.
Supported layers
nn.Linearnn.ReLUnn.Sequentialnn.Sigmoidnn.Softplusnn.Tanh
Keyword arguments
config: a dictionary that maps Symbols to AbstractPredictors that control how the activation functions are reformulated. For example, :Sigmoid => MathOptAI.Sigmoid() or :ReLU => MathOptAI.QuadraticReLU(). The supported Symbols are :ReLU, :Sigmoid, :SoftPlus, and :Tanh.gray_box: if true, the neural network is added as a user-defined nonlinear operator, with gradients provided by torch.func.jacrev.gray_box_hessian: if true, the gray box additionally computes the Hessian of the output using torch.func.hessian.gray_box_device: device used to construct PyTorch tensors, e.g. "cuda" to run on an Nvidia GPU.
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
+)
Convert a trained neural network from PyTorch via PythonCall.jl to a Pipeline.
Supported layers
nn.Linearnn.ReLUnn.Sequentialnn.Sigmoidnn.Softplusnn.Tanh
Keyword arguments
config: a dictionary that maps Symbols to AbstractPredictors that control how the activation functions are reformulated. For example, :Sigmoid => MathOptAI.Sigmoid() or :ReLU => MathOptAI.QuadraticReLU(). The supported Symbols are :ReLU, :Sigmoid, :SoftPlus, and :Tanh.gray_box: if true, the neural network is added as a user-defined nonlinear operator, with gradients provided by torch.func.jacrev.gray_box_hessian: if true, the gray box additionally computes the Hessian of the output using torch.func.hessian.gray_box_device: device used to construct PyTorch tensors, e.g. "cuda" to run on an Nvidia GPU.
sourceMathOptAI.add_predictor — MethodMathOptAI.add_predictor(
model::JuMP.AbstractModel,
predictor::StatsModels.TableRegressionModel,
x::DataFrames.DataFrame;
@@ -891,4 +891,4 @@
moai_Affine[1]
moai_Affine[1]
moai_Affine[1]
- moai_Affine[1]
sourceSettings
This document was generated with Documenter.jl version 1.7.0 on Thursday 7 November 2024. Using Julia version 1.11.1.