Thanks to visit codestin.com
Credit goes to github.com

Skip to content
This repository was archived by the owner on Jul 1, 2023. It is now read-only.

Revert non-active try-apply differentiation workarounds #1069

Merged
merged 1 commit into from
Dec 4, 2020
Merged

Revert non-active try-apply differentiation workarounds #1069

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
45 changes: 15 additions & 30 deletions Sources/TensorFlow/Operators/Math.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1846,8 +1846,7 @@ extension Tensor where Scalar: Numeric & Comparable {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func max(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return max(squeezingAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -1876,8 +1875,7 @@ extension Tensor where Scalar: Numeric & Comparable {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func min(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return min(squeezingAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -1928,8 +1926,7 @@ extension Tensor where Scalar: Numeric & Comparable {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func min(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return min(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -1961,8 +1958,7 @@ extension Tensor where Scalar: Numeric & Comparable {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func max(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return max(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2100,8 +2096,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func sum(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int64.init) }()
let axes = axes.map(Int64.init)
return _Raw.sum(self, reductionIndices: axes, keepDims: false)
}

Expand Down Expand Up @@ -2136,8 +2131,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func sum(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int64.init) }()
let axes = axes.map(Int64.init)
return _Raw.sum(self, reductionIndices: axes, keepDims: true)
}

Expand Down Expand Up @@ -2170,8 +2164,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func product(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return product(squeezingAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2207,8 +2200,7 @@ extension Tensor where Scalar: Numeric {
/// - Precondition: Each value in `axes` must be in the range `-rank..<rank`.
@inlinable
public func product(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return product(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2239,8 +2231,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func mean(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int64.init) }()
let axes = axes.map(Int64.init)
return _Raw.mean(self, reductionIndices: axes, keepDims: false)
}

Expand Down Expand Up @@ -2277,8 +2268,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func mean(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int64.init) }()
let axes = axes.map(Int64.init)
return _Raw.mean(self, reductionIndices: axes, keepDims: true)
}

Expand Down Expand Up @@ -2313,8 +2303,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func variance(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return variance(squeezingAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2355,8 +2344,7 @@ extension Tensor where Scalar: Numeric {
@inlinable
@differentiable(wrt: self where Scalar: TensorFlowFloatingPoint)
public func variance(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return variance(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2777,8 +2765,7 @@ extension Tensor where Scalar: TensorFlowFloatingPoint {
@inlinable
@differentiable(wrt: self)
public func standardDeviation(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = { axes.map(Int32.init) }()
let axes = axes.map(Int32.init)
return standardDeviation(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2828,8 +2815,7 @@ extension Tensor where Scalar: TensorFlowFloatingPoint {
@inlinable
@differentiable(wrt: self)
public func logSumExp(squeezingAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = withoutDerivative(at: axes) { $0.map(Int32.init) }
let axes = axes.map(Int32.init)
return logSumExp(squeezingAxes: Tensor<Int32>(axes, on: device))
}

Expand Down Expand Up @@ -2893,8 +2879,7 @@ extension Tensor where Scalar: TensorFlowFloatingPoint {
@inlinable
@differentiable(wrt: self)
public func logSumExp(alongAxes axes: [Int]) -> Tensor {
// TODO(TF-433): Remove workaround for differentiating `map`.
let axes = withoutDerivative(at: axes) { $0.map(Int32.init) }
let axes = axes.map(Int32.init)
return logSumExp(alongAxes: Tensor<Int32>(axes, on: device))
}

Expand Down