An options class for controlling runtime interpreter behavior.
Nested Classes
Public Constructors
| * | Options
()
|
| * |
Public Methods
| InterpreterApi.Options | |
| InterpreterApi.Options | addDelegateFactory
( DelegateFactory
delegateFactory)
Adds a
DelegateFactory
which will be invoked to apply its created Delegate
during interpreter creation. |
| ValidatedAccelerationConfig | |
| List < DelegateFactory > | getDelegateFactories
()
Returns the list of delegate factories that have been registered via
addDelegateFactory
). |
| List < Delegate > | getDelegates
()
Returns the list of delegates intended to be applied during interpreter creation that have
been registered via
addDelegate
. |
| int | |
| InterpreterApi.Options.TfLiteRuntime | |
| boolean | |
| boolean | |
| boolean | |
| InterpreterApi.Options | |
| InterpreterApi.Options | |
| InterpreterApi.Options | setNumThreads
(int numThreads)
Sets the number of threads to be used for ops that support multi-threading.
|
| InterpreterApi.Options | setRuntime
( InterpreterApi.Options.TfLiteRuntime
runtime)
Specify where to get the TF Lite runtime implementation from.
|
| InterpreterApi.Options | |
| InterpreterApi.Options | setUseXNNPACK
(boolean useXNNPACK)
Enable or disable an optimized set of CPU kernels (provided by XNNPACK).
|
Inherited Methods
Public Constructors
public Options ()
public Options ( InterpreterApi.Options other)
Parameters
Public Methods
public InterpreterApi.Options addDelegate ( Delegate delegate)
Adds a Delegate
to be applied during interpreter creation.
Delegates added here are applied before any delegates created from a DelegateFactory
that was added with addDelegateFactory(DelegateFactory)
.
Note that TF Lite in Google Play Services (see setRuntime(InterpreterApi.Options.TfLiteRuntime)
) does not support
external (developer-provided) delegates, and adding a Delegate
other than ERROR(/NnApiDelegate)
here is not allowed when using TF Lite in Google Play Services.
Parameters
public InterpreterApi.Options addDelegateFactory ( DelegateFactory delegateFactory)
Adds a DelegateFactory
which will be invoked to apply its created Delegate
during interpreter creation.
Delegates from a delegated factory that was added here are applied after any delegates
added with addDelegate(Delegate)
.
Parameters
public ValidatedAccelerationConfig getAccelerationConfig ()
Return the acceleration configuration.
public List < DelegateFactory > getDelegateFactories ()
Returns the list of delegate factories that have been registered via addDelegateFactory
).
public List < Delegate > getDelegates ()
Returns the list of delegates intended to be applied during interpreter creation that have
been registered via addDelegate
.
public int getNumThreads ()
Returns the number of threads to be used for ops that support multi-threading.
numThreads
should be >= -1
. Values of 0 (or 1) disable multithreading.
Default value is -1: the number of threads used will be implementation-defined and
platform-dependent.
public InterpreterApi.Options.TfLiteRuntime getRuntime ()
Return where to get the TF Lite runtime implementation from.
public boolean getUseNNAPI ()
Returns whether to use NN API (if available) for op execution. Default value is false (disabled).
public boolean getUseXNNPACK ()
public boolean isCancellable ()
Advanced: Returns whether the interpreter is able to be cancelled.
Interpreters may have an experimental API setCancelled(boolean)
.
If this interpreter is cancellable and such a method is invoked, a cancellation flag will be
set to true. The interpreter will check the flag between Op invocations, and if it's true
, the interpreter will stop execution. The interpreter will remain a cancelled state
until explicitly "uncancelled" by setCancelled(false)
.
public InterpreterApi.Options setAccelerationConfig ( ValidatedAccelerationConfig config)
Specify the acceleration configuration.
Parameters
public InterpreterApi.Options setCancellable (boolean allow)
Advanced: Set if the interpreter is able to be cancelled.
Interpreters may have an experimental API setCancelled(boolean)
.
If this interpreter is cancellable and such a method is invoked, a cancellation flag will be
set to true. The interpreter will check the flag between Op invocations, and if it's true
, the interpreter will stop execution. The interpreter will remain a cancelled state
until explicitly "uncancelled" by setCancelled(false)
.
Parameters
public InterpreterApi.Options setNumThreads (int numThreads)
Sets the number of threads to be used for ops that support multi-threading.
numThreads
should be >= -1
. Setting numThreads
to 0 has the effect
of disabling multithreading, which is equivalent to setting numThreads
to 1. If
unspecified, or set to the value -1, the number of threads used will be
implementation-defined and platform-dependent.
Parameters
public InterpreterApi.Options setRuntime ( InterpreterApi.Options.TfLiteRuntime runtime)
Specify where to get the TF Lite runtime implementation from.
Parameters
public InterpreterApi.Options setUseNNAPI (boolean useNNAPI)
Sets whether to use NN API (if available) for op execution. Defaults to false (disabled).
Parameters
public InterpreterApi.Options setUseXNNPACK (boolean useXNNPACK)
Enable or disable an optimized set of CPU kernels (provided by XNNPACK). Enabled by default.
