Introducing Google AI Edge Portal: Benchmark Edge AI at scale. Sign-up to request access during private preview.

Class Index

These are all the API classes. See all API packages.

A B C D E I J L M N R S T

A

`Accelerator`	Hardware accelerators supported by LiteRT.
`AiPackModelProvider`	A `ModelProvider` that's backed by on-demand Google Play AI Pack.

BuiltinNpuAcceleratorProvider

An implementation of NpuAcceleratorProvider, which provides the NPU libraries without dynamic downloading.

`CompiledModel`	Class that represents a compiled LiteRT model.
`CompiledModel.CpuOptions`	Options to specify CPU acceleration for compiling a model.
`CompiledModel.GpuOptions`	Options to specify GPU acceleration for compiling a model.
`CompiledModel.GpuOptions.Backend`	Backend for GPU options.
`CompiledModel.GpuOptions.BufferStorageType`	Buffer storage type for GPU options.
`CompiledModel.GpuOptions.Precision`	Precision for GPU options.
`CompiledModel.GpuOptions.Priority`	Priority for GPU options.
`CompiledModel.Options`	Options to specify hardware acceleration for compiling a model.
`CompiledModel.QualcommOptions`	Options to specify Qualcomm options for compiling a model.
`CompiledModel.QualcommOptions.HtpPerformanceMode`	HTP performance mode for Qualcomm options.
`CompiledModel.QualcommOptions.LogLevel`	Log level for Qualcomm options.
`CompiledModel.QualcommOptions.OptimizationLevel`	Optimization level for Qualcomm options.
`CompiledModel.QualcommOptions.Profiling`	Profiling for Qualcomm options.

DataType

Represents the type of elements in a TensorFlow Lite Tensor as an enum.

`Environment`	Environment to hold configuration options for LiteRT runtime.
`Environment.Option`	Options configurable in LiteRT environment.

`InterpreterApi`	Interface to TensorFlow Lite model interpreter, excluding experimental methods.
`InterpreterApi.Options`	An options class for controlling runtime interpreter behavior.
`InterpreterApi.Options.TfLiteRuntime`	Enum to represent where to get the TensorFlow Lite runtime implementation from.
`InterpreterFactory`	Factory for constructing InterpreterApi instances.

JniHandle

A base class for all Kotlin types that wrap a JNI handle.

LiteRtException

Exception for various LiteRT API errors.

`ModelProvider`	A model provider that provides a model file and relevant information.
`ModelProvider.Type`	Model files could be either an asset or a file.
`ModelSelector`	ModelSelector allows to dynamically select a `ModelProvider` from a given set of providers.

`NpuAcceleratorProvider`	An interface to provide the NPU libraries.
`NpuCompatibilityChecker`	An interface to checks if the device is compatible with NPU.

RuntimeFlavor

Represents a TFLite runtime.

Status

Status codes for LiteRtException.

`Tensor`	A typed multi-dimensional array used in Tensorflow Lite.
`Tensor.QuantizationParams`	Quantization parameters that corresponds to the table, `QuantizationParameters`, in the TFLite Model schema file.
`TensorBuffer`	TensorBuffer represents the raw memory where tensor data is stored.
`TensorBufferRequirements`	Requirements for allocating a TensorBuffer.
`TensorBufferType`	The type of the tensor buffer.
`TensorFlowLite`	Static utility methods for loading the TensorFlowLite runtime.
`TensorType`	The type of a tensor, including its element type and layout.
`TensorType.ElementType`	Data type of tensor elements.
`TensorType.Layout`	Layout of a tensor.