History

YoungSeok Yoon 772433a2a2 Add flag for using optimized TFLite CPU kernels on iOS This adds new experimental flags to the interpreter options of TFLite Obj-C and Swift APIs, which can be used for opting in to a set of highly optimized floating point kernels provided via the XNNPACK delegate. The flags can be used as follows. Obj-C: TFLInterpreterOptions options = [[TFLInterpreterOptions alloc] init]; options.useXNNPACK = YES; NSError error; TFLInterpreter *interpreter = [[TFLInterpreter alloc] initWithModelPath:@"model/path" options:options error:&error]; Swift: var options = InterpreterOptions() options.isXNNPackEnabled = true var interpreter = try Interpreter(modelPath: "model/path", options: options) PiperOrigin-RevId: 317270012 Change-Id: I82aae43c3de13ab08af3c70513e2a458e807b0f1		2020-06-19 02:03:48 -07:00
..
Sources	Add flag for using optimized TFLite CPU kernels on iOS	2020-06-19 02:03:48 -07:00
TensorFlowLite.tulsiproj	Clean up the Swift TestApp and direct users to TFLite example apps.	2020-06-07 17:19:10 -07:00
TestApp	Clean up the Swift TestApp and direct users to TFLite example apps.	2020-06-07 17:19:10 -07:00
Tests	Add flag for using optimized TFLite CPU kernels on iOS	2020-06-19 02:03:48 -07:00
BUILD.apple	Add build-related tags to TensorFlowLite and TensorFlowLiteAllDelegates	2020-06-08 21:45:02 -07:00
README.md	Adds the TFLite version string to the Swift library.	2019-05-17 11:29:21 -07:00
TensorFlowLiteSwift-nightly.podspec	Generate separate pod for Core ML delegate	2020-05-14 19:07:57 -07:00
TensorFlowLiteSwift.podspec	Podspec updates for TFLite v2.2.0	2020-05-06 21:50:18 -07:00
TensorFlowLiteSwift.podspec.template	Move test_spec in TensorFlowLiteSwift.podspec.template into subspec, so that the Delegate tests are only done in corresponding delegates.	2020-05-26 21:04:08 -07:00

README.md

TensorFlow Lite for Swift

TensorFlow Lite is TensorFlow's lightweight solution for Swift developers. It enables low-latency inference of on-device machine learning models with a small binary size and fast performance supporting hardware acceleration.

Build TensorFlow with iOS support

To build the Swift TensorFlow Lite library on Apple platforms, install from source or clone the GitHub repo. Then, configure TensorFlow by navigating to the root directory and executing the configure.py script:

python configure.py

Follow the prompts and when asked to build TensorFlow with iOS support, enter y.

CocoaPods developers

Add the TensorFlow Lite pod to your Podfile:

pod 'TensorFlowLiteSwift'

Then, run pod install.

In your Swift files, import the module:

import TensorFlowLite

Bazel developers

In your BUILD file, add the TensorFlowLite dependency to your target:

swift_library(
  deps = [
      "//tensorflow/lite/experimental/swift:TensorFlowLite",
  ],
)

In your Swift files, import the module:

import TensorFlowLite

Build the TensorFlowLite Swift library target:

bazel build tensorflow/lite/experimental/swift:TensorFlowLite

Build the Tests target:

bazel test tensorflow/lite/experimental/swift:Tests --swiftcopt=-enable-testing

Note: --swiftcopt=-enable-testing is required for optimized builds (-c opt).

Generate the Xcode project using Tulsi

Open the //tensorflow/lite/experimental/swift/TensorFlowLite.tulsiproj using the TulsiApp or by running the generate_xcodeproj.sh script from the root tensorflow directory:

generate_xcodeproj.sh --genconfig tensorflow/lite/experimental/swift/TensorFlowLite.tulsiproj:TensorFlowLite --outputfolder ~/path/to/generated/TensorFlowLite.xcodeproj