Class/Object

ml.dmlc.mxnet

Executor

Related Docs: object Executor | package mxnet

Permalink

class Executor extends AnyRef

Symbolic Executor component of MXNet
WARNING: it is your responsibility to clear this object through dispose(). NEVER rely on the GC strategy

See also

Symbol.bind : to create executor

Linear Supertypes
AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. Executor
  2. AnyRef
  3. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Value Members

  1. final def !=(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  2. final def ##(): Int

    Permalink
    Definition Classes
    AnyRef → Any
  3. final def ==(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  4. var _argDict: Map[String, NDArray]

    Permalink
    Attributes
    protected
  5. var _auxDict: Map[String, NDArray]

    Permalink
    Attributes
    protected
  6. var _gradDict: Map[String, NDArray]

    Permalink
    Attributes
    protected
  7. def argDict: Map[String, NDArray]

    Permalink

    Get dictionary representation of argument arrrays.

    Get dictionary representation of argument arrrays.

    returns

    The dictionary that maps name of arguments to NDArrays.

    Exceptions thrown

    IllegalArgumentException if there are duplicated names in the arguments.

  8. final def asInstanceOf[T0]: T0

    Permalink
    Definition Classes
    Any
  9. def auxDict: Map[String, NDArray]

    Permalink

    Get dictionary representation of auxiliary states arrays.

    Get dictionary representation of auxiliary states arrays.

    returns

    The dictionary that maps name of auxiliary states to NDArrays.

    Exceptions thrown

    IllegalArgumentException if there are duplicated names in the auxiliary states.

  10. def backward(): Unit

    Permalink
  11. def backward(outGrad: NDArray): Unit

    Permalink
  12. def backward(outGrads: Array[NDArray]): Unit

    Permalink

    Do backward pass to get the gradient of arguments.

    Do backward pass to get the gradient of arguments.

    outGrads

    Gradient on the outputs to be propagated back. This parameter is only needed when bind is called on outputs that are not a loss function.

  13. def clone(): AnyRef

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  14. def copyParamsFrom(argParams: Map[String, NDArray]): Unit

    Permalink
  15. def copyParamsFrom(argParams: Map[String, NDArray], allowExtraParams: Boolean): Unit

    Permalink
  16. def copyParamsFrom(argParams: Map[String, NDArray], auxParams: Map[String, NDArray], allowExtraParams: Boolean = false): Unit

    Permalink

    Copy parameters from arg_params, aux_params into executor's internal array.

    Copy parameters from arg_params, aux_params into executor's internal array.

    argParams

    : dict of name to NDArray of arguments

    auxParams

    : dict of name to NDArray of auxiliary states.

    allowExtraParams

    Whether allow extra parameters that are not needed by symbol If this is True, no error will be thrown when arg_params or aux_params contain extra parameters that is not needed by the executor.

    Exceptions thrown

    IllegalArgumentException If there is additional parameters in the dict but allow_extra_params=False

  17. def debugStr: String

    Permalink

    Get a debug string about internal execution plan.

    Get a debug string about internal execution plan.

    returns

    Debug string of the executor.

  18. def dispose(): Unit

    Permalink
  19. final def eq(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  20. def equals(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  21. def finalize(): Unit

    Permalink
    Attributes
    protected
    Definition Classes
    Executor → AnyRef
  22. def forward(): Unit

    Permalink
  23. def forward(isTrain: Boolean, kwargs: (String, NDArray)*): Unit

    Permalink

    Calculate the outputs specified by the binded symbol.

    Calculate the outputs specified by the binded symbol.

    isTrain

    whether this forward is for evaluation purpose.

    kwargs

    Additional specification of input arguments.

  24. final def getClass(): Class[_]

    Permalink
    Definition Classes
    AnyRef → Any
  25. def gradDict: Map[String, NDArray]

    Permalink

    Get dictionary representation of gradient arrays.

    Get dictionary representation of gradient arrays.

    returns

    The dictionary that maps name of arguments to gradient arrays.

    Exceptions thrown

    IllegalArgumentException if there are duplicated names in the grads.

  26. def hashCode(): Int

    Permalink
    Definition Classes
    AnyRef → Any
  27. final def isInstanceOf[T0]: Boolean

    Permalink
    Definition Classes
    Any
  28. var monitorCallback: MXMonitorCallback

    Permalink
    Attributes
    protected
  29. final def ne(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  30. final def notify(): Unit

    Permalink
    Definition Classes
    AnyRef
  31. final def notifyAll(): Unit

    Permalink
    Definition Classes
    AnyRef
  32. val outputs: Array[NDArray]

    Permalink
  33. def reshape(partialShaping: Boolean = false, allowUpSizing: Boolean = false, kwargs: Map[String, Shape]): Executor

    Permalink

    Return a new executor with the same symbol and shared memory, but different input/output shapes.

    Return a new executor with the same symbol and shared memory, but different input/output shapes. For runtime reshaping, variable length sequences, etc. The returned executor shares state with the current one, and cannot be used in parallel with it.

    partialShaping

    Whether to allow changing the shape of unspecified arguments.

    allowUpSizing

    Whether to allow allocating new ndarrays that's larger than the original.

    kwargs

    Map of string to Shape.

    • new shape for arguments.
    returns

    executor A new executor that shares memory with this.

  34. def setMonitorCallback(callback: MXMonitorCallback): Unit

    Permalink

    Install callback.

    Install callback.

    callback

    Takes a string and an NDArrayHandle.

  35. final def synchronized[T0](arg0: ⇒ T0): T0

    Permalink
    Definition Classes
    AnyRef
  36. def toString(): String

    Permalink
    Definition Classes
    AnyRef → Any
  37. final def wait(): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  38. final def wait(arg0: Long, arg1: Int): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  39. final def wait(arg0: Long): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )

Inherited from AnyRef

Inherited from Any

Ungrouped