Swift for TensorFlow

E97037
differentiable programming platform experimental software project machine learning framework open-source project

Swift for TensorFlow is an experimental machine learning platform that integrates TensorFlow directly into the Swift programming language to enable differentiable programming and high-performance model development.

Jump to: Surface forms Statements Referenced by

Observed surface forms (1)

Surface form Occurrences
TensorFlow runtime bindings 1

Statements (49)

Predicate Object
instanceOf differentiable programming platform ⓘ
experimental software project ⓘ
machine learning framework ⓘ
open-source project ⓘ
aimsTo enable high-performance machine learning ⓘ
integrate machine learning into general-purpose programming ⓘ
provide first-class differentiable programming in Swift ⓘ
announcedBy Chris Lattner ⓘ
basedOn Swift ⓘ
developedBy Google ⓘ
TensorFlow ⓘ
surface form: TensorFlow team
discontinued true ⓘ
enables compiling models for accelerators ⓘ
training neural networks in Swift ⓘ
writing differentiable Swift functions ⓘ
hasComponent Swift compiler modifications ⓘ
Swift for TensorFlow self-linksurface differs ⓘ
surface form: TensorFlow runtime bindings

model training APIs ⓘ
standard library additions ⓘ
integratesWith TensorFlow ⓘ
license Apache License 2.0 ⓘ
partOf TensorFlow ecosystem ⓘ
programmingLanguage Swift ⓘ
relatedTo JAX ⓘ
PyTorch ⓘ
TensorFlow ⓘ
repository https://github.com/tensorflow/swift ⓘ
status archived ⓘ
experimental ⓘ
supportsDomain deep learning ⓘ
numerical computing ⓘ
supportsFeature GPU acceleration ⓘ
Python interoperability ⓘ
TPU acceleration ⓘ
automatic differentiation ⓘ
custom differentiable types ⓘ
differentiable programming ⓘ
eager execution ⓘ
first-class gradients ⓘ
graph execution ⓘ
model inference ⓘ
model training ⓘ
supportsLanguage Swift ⓘ
targetUser machine learning engineers ⓘ
machine learning researchers ⓘ
systems programmers ⓘ
usesConcept protocol-oriented programming ⓘ
static typing ⓘ
value semantics ⓘ

Referenced by (2)

Full triples — surface form annotated when it differs from this entity's canonical label.

Swift for TensorFlow → hasComponent → Swift for TensorFlow self-linksurface differs ⓘ
this entity surface form: TensorFlow runtime bindings
Swift → influenced → Swift for TensorFlow ⓘ
subject surface form: Swift (programming language)