swift-apis
swift-apis copied to clipboard
tensorflow
Swift for TensorFlow Deep Learning Library
ShapedArray is a very use-full and power tool, and would like to be able to install it separately
Currently this builds in `Debug` on the latest nightly toolchain on Linux. macOS is untested. All unit tests pass except one, which is not crucial (only a floating point difference...
Opening [a blank notebook](https://colab.research.google.com/github/tensorflow/swift/blob/master/notebooks/blank_swift.ipynb#scrollTo=kZRlD4utdPuX) which I expect is running S4TF v0.9 and entering the following: import TensorFlow import x10_training_loop Device.trainingDevices // Same error for runOnThreads(), HostStatistics(), ... etc. Gives the...
Add GitHub Actions workflow for: - Downloading and installing a swift.org/download macOS development snapshot - Building tensorflow/swift-apis via CMake - Installing tensorflow/swift-apis into the toolchain via CMake - Repackaging the...
`VectorProtocol.VectorSpaceScalar` [was changed](https://github.com/tensorflow/swift-apis/pull/1139/files#r537911457) from an associated type to a hardcoded typealias for `Float`. (I believe this simplification (removing `VectorSpaceScalar` as a customization point) made it easier to enable building `tensorflow/swift-apis`...
Now that we support a multitude of basic layers, I'd like this issue to serve as the discussion for supporting advanced layers. While #54 tracked basic layers and most have...
[Here](https://github.com/tensorflow/swift-apis/blob/b7148bc6f5e76bd55e5b11ebd5dd14e42c6ef5a3/Sources/TensorFlow/Core/ShapedArray.swift#L269) is `ShapedArray`'s ` fileprivate func description( indentLevel: Int, edgeElementCount: Int, maxScalarLength: Int, maxScalarCountPerLine: Int, summarizing: Bool ) -> String`. Is there any reason this is marked `fileprivate`? It's currently...
Associated type inference behavior was changed in [apple/swift#32578](https://github.com/apple/swift/pull/32578): derived conformances are now attempted before associated type inference. This broke `ParameterlessLayer`, which relied on a `TangentVector == EmptyTangentVector` same-type constraint to...
On a machine with GPU or TPU, I get a segfault if I try to use `Device` with CPU type on XLA backend, e.g.: ``` let device = Device(kind: .CPU,...
The Tensor initializer lets you manually specify a device for placement. For example, you should be able to explicitly place a Tensor on the first CPU using the eager mode...
