Why Go
- Case Studies
  
  Common problems companies solve with Go
- Use Cases
  
  Stories about how and why companies use Go
- Security
  
  How Go can help keep you secure by default
Learn
Docs
- Effective Go
  
  Tips for writing clear, performant, and idiomatic Go code
- Go User Manual
  
  A complete introduction to building software with Go
- Standard library
  
  Reference documentation for Go's standard library
- Release Notes
  
  Learn what's new in each Go release
Packages
Community
- Recorded Talks
  
  Videos from prior events
- Meetups
  
  Meet other local Go developers
- Conferences
  
  Learn and network with Go developers from around the world
- Go blog
  
  The Go project's official blog.
- Go project
  
  Get help and stay informed from Go
- Get connected

compute

package

Go to main page

Versions in this module

v0

v0.3.0

Aug 25, 2025

Changes in this version

type CPUEngine

+ func (e *CPUEngine[T]) ReduceMean(ctx context.Context, a *tensor.TensorNumeric[T], axis int, keepDims bool, ...) (*tensor.TensorNumeric[T], error)

+ func (e *CPUEngine[T]) Rsqrt(ctx context.Context, a *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

+ func (e *CPUEngine[T]) Tanh(ctx context.Context, a *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

+ func (e *CPUEngine[T]) TanhPrime(ctx context.Context, a, upstream *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

type Engine

+ ReduceMean func(ctx context.Context, a *tensor.TensorNumeric[T], axis int, keepDims bool, ...) (*tensor.TensorNumeric[T], error)

+ Rsqrt func(ctx context.Context, a *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

+ Tanh func(ctx context.Context, a *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

+ TanhPrime func(ctx context.Context, a, upstream *tensor.TensorNumeric[T], ...) (*tensor.TensorNumeric[T], error)

v0.1.0

Aug 4, 2025

Changes in this version

+ type CPUEngine struct

+ func NewCPUEngine[T tensor.Numeric](ops numeric.Arithmetic[T]) *CPUEngine[T]

+ func (e *CPUEngine[T]) Add(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) AddScalar(_ context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Concat(_ context.Context, tensors []*tensor.Tensor[T], axis int, ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Copy(_ context.Context, dst, src *tensor.Tensor[T]) error

+ func (e *CPUEngine[T]) Div(_ context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) DivScalar(_ context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Exp(_ context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Fill(_ context.Context, t *tensor.Tensor[T], value T) error

+ func (e *CPUEngine[T]) Gather(_ context.Context, params *tensor.Tensor[T], indices *tensor.Tensor[int], ...) error

+ func (e *CPUEngine[T]) Log(_ context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) MatMul(_ context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Mul(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) MulScalar(_ context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) OneHot(_ context.Context, input *tensor.Tensor[int], depth int, ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Ops() numeric.Arithmetic[T]

+ func (e *CPUEngine[T]) Pow(ctx context.Context, base, exponent *tensor.Tensor[T], ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) RandomUniform(_ context.Context, t *tensor.Tensor[T], minVal, maxVal T) error

+ func (e *CPUEngine[T]) ReduceSum(ctx context.Context, a *tensor.Tensor[T], axis int, keepDims bool, ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Repeat(_ context.Context, a *tensor.Tensor[T], axis int, repetitions int, ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Reshape(_ context.Context, a *tensor.Tensor[T], shape []int, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) ScatterAdd(_ context.Context, dEmbeddingTable *tensor.Tensor[T], ...) error

+ func (e *CPUEngine[T]) Softmax(_ context.Context, a *tensor.Tensor[T], axis int, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Split(_ context.Context, a *tensor.Tensor[T], numSplits int, axis int) ([]*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Sqrt(_ context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Sub(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Sum(ctx context.Context, a *tensor.Tensor[T], axis int, keepDims bool, ...) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Transpose(_ context.Context, a *tensor.Tensor[T], axes []int, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) UnaryOp(_ context.Context, a *tensor.Tensor[T], op func(T) T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ func (e *CPUEngine[T]) Zero(_ context.Context, a *tensor.Tensor[T]) error

+ func (e *CPUEngine[T]) Zeros(_ context.Context, a *tensor.Tensor[T], shape []int) error

+ type Engine interface

+ Add func(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ AddScalar func(ctx context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Concat func(ctx context.Context, tensors []*tensor.Tensor[T], axis int, ...) (*tensor.Tensor[T], error)

+ Copy func(ctx context.Context, dst, src *tensor.Tensor[T]) error

+ Div func(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ DivScalar func(ctx context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Exp func(ctx context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Fill func(ctx context.Context, t *tensor.Tensor[T], value T) error

+ Gather func(ctx context.Context, params *tensor.Tensor[T], indices *tensor.Tensor[int], ...) error

+ Log func(ctx context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ MatMul func(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Mul func(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ MulScalar func(ctx context.Context, a *tensor.Tensor[T], scalar T, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ OneHot func(ctx context.Context, input *tensor.Tensor[int], depth int, ...) (*tensor.Tensor[T], error)

+ Ops func() numeric.Arithmetic[T]

+ Pow func(ctx context.Context, base, exponent *tensor.Tensor[T], ...) (*tensor.Tensor[T], error)

+ RandomUniform func(ctx context.Context, t *tensor.Tensor[T], minVal, maxVal T) error

+ ReduceSum func(ctx context.Context, a *tensor.Tensor[T], axis int, keepDims bool, ...) (*tensor.Tensor[T], error)

+ Repeat func(ctx context.Context, a *tensor.Tensor[T], axis int, repetitions int, ...) (*tensor.Tensor[T], error)

+ Reshape func(ctx context.Context, a *tensor.Tensor[T], shape []int, ...) (*tensor.Tensor[T], error)

+ ScatterAdd func(ctx context.Context, dEmbeddingTable *tensor.Tensor[T], ...) error

+ Softmax func(ctx context.Context, a *tensor.Tensor[T], axis int, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Split func(ctx context.Context, a *tensor.Tensor[T], numSplits int, axis int) ([]*tensor.Tensor[T], error)

+ Sqrt func(ctx context.Context, a *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Sub func(ctx context.Context, a, b *tensor.Tensor[T], dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ Sum func(ctx context.Context, a *tensor.Tensor[T], axis int, keepDims bool, ...) (*tensor.Tensor[T], error)

+ Transpose func(ctx context.Context, a *tensor.Tensor[T], axes []int, dst ...*tensor.Tensor[T]) (*tensor.Tensor[T], error)

+ UnaryOp func(ctx context.Context, a *tensor.Tensor[T], op func(T) T, ...) (*tensor.Tensor[T], error)

+ Zero func(ctx context.Context, a *tensor.Tensor[T]) error

+ Zeros func(ctx context.Context, a *tensor.Tensor[T], shape []int) error

+ type FailableTensor struct

+ func NewFailableTensor[T tensor.Numeric](t *tensor.Tensor[T]) *FailableTensor[T]

+ func (f *FailableTensor[T]) Set(value T, indices ...int) error

+ func (f *FailableTensor[T]) SetFailOnSet(fail bool)

+ func (f *FailableTensor[T]) SetFailOnSetAfter(count int)

+ type FailableZeroer struct

+ func NewFailableZeroer[T tensor.Numeric](engine *TestableEngine[T]) *FailableZeroer[T]

+ func (f *FailableZeroer[T]) SetFailOnZero(fail bool)

+ func (f *FailableZeroer[T]) Zero(ctx context.Context, a *tensor.Tensor[T]) error

+ type TestableEngine struct

+ func NewTestableEngine[T tensor.Numeric](ops numeric.Arithmetic[T]) *TestableEngine[T]

+ func (e *TestableEngine[T]) TestableMatMul(_ context.Context, a, b *tensor.Tensor[T], result *FailableTensor[T]) error

+ func (e *TestableEngine[T]) TestableSum(ctx context.Context, a *tensor.Tensor[T], axis int, _ bool, ...) error

+ func (e *TestableEngine[T]) TestableTranspose(_ context.Context, a *tensor.Tensor[T], result *FailableTensor[T]) error