Class/Object

ml.dmlc.mxnet

FeedForward

Related Docs: object FeedForward | package mxnet

Permalink

class FeedForward extends AnyRef

Model class of MXNet for training and predicting feedforward nets. This class is designed for a single-data single output supervised network.

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

Instance Constructors

  1. new FeedForward(symbol: SymbolGenerator, ctx: Array[Context], numEpoch: Int, epochSize: Int, optimizer: Optimizer, initializer: Initializer, batchSize: Int, argParams: Map[String, NDArray], auxParams: Map[String, NDArray], allowExtraParams: Boolean, beginEpoch: Int)

    Permalink
  2. new FeedForward(symbol: Symbol, ctx: Array[Context] = Array(Context.cpu()), numEpoch: Int = 1, epochSize: Int = 1, optimizer: Optimizer = new SGD(), initializer: Initializer = new Uniform(0.01f), batchSize: Int = 128, argParams: Map[String, NDArray] = null, auxParams: Map[String, NDArray] = null, allowExtraParams: Boolean = false, beginEpoch: Int = 0)

    Permalink

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. final def asInstanceOf[T0]: T0

    Permalink
    Definition Classes
    Any
  5. val beginEpoch: Int

    Permalink

    The beginning training epoch.

  6. def checkArguments(): Unit

    Permalink
  7. def clone(): AnyRef

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  8. val epochSize: Int

    Permalink

    Number of batches in a epoch.

    Number of batches in a epoch. In default, it is set to ceil(num_train_examples / batch_size)

  9. final def eq(arg0: AnyRef): Boolean

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

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

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( classOf[java.lang.Throwable] )
  12. def fit(trainData: DataIter, evalData: DataIter, kvStore: KVStore): Unit

    Permalink
  13. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kvStore: KVStore): Unit

    Permalink
  14. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kvStore: KVStore, epochEndCallback: EpochEndCallback, batchEndCallback: BatchEndCallback): Unit

    Permalink
  15. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kv: KVStore, epochEndCallback: EpochEndCallback, batchEndCallback: BatchEndCallback, logger: <error>, workLoadList: Seq[Float]): Unit

    Permalink
  16. def fit(trainData: DataIter, evalData: DataIter): Unit

    Permalink
  17. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric): Unit

    Permalink
  18. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kvStoreType: String): Unit

    Permalink
  19. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kvStoreType: String, epochEndCallback: EpochEndCallback, batchEndCallback: BatchEndCallback): Unit

    Permalink
  20. def fit(trainData: DataIter, evalData: DataIter, evalMetric: EvalMetric, kvStoreType: String, epochEndCallback: EpochEndCallback, batchEndCallback: BatchEndCallback, logger: <error>, workLoadList: Seq[Float]): Unit

    Permalink

    Fit the model.

    Fit the model.

    trainData

    Training data

    evalData

    Evaluation data

    evalMetric

    The evaluation metric, cannot be null

    kvStoreType

    A string kvstore type: 'local' : multi-devices on a single machine, will automatically choose one from 'local_update_cpu', 'local_allreduce_cpu', and 'local_allreduce_device' 'dist_sync' : multi-machines with BSP 'dist_async' : multi-machines with partical asynchronous In default uses 'local', often no need to change for single machine.

    epochEndCallback

    A callback that is invoked at end of each epoch. This can be used to checkpoint model each epoch.

    batchEndCallback

    A callback that is invoked at end of each batch For print purpose

    logger

    When not specified, default logger will be used.

    workLoadList

    The list of work load for different devices, in the same order as ctx

  21. def getArgParams: Map[String, NDArray]

    Permalink
  22. def getAuxParams: Map[String, NDArray]

    Permalink
  23. final def getClass(): Class[_]

    Permalink
    Definition Classes
    AnyRef → Any
  24. def hashCode(): Int

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

    Permalink
    Definition Classes
    Any
  26. val logger: <error>

    Permalink
  27. final def ne(arg0: AnyRef): Boolean

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

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

    Permalink
    Definition Classes
    AnyRef
  30. var predExec: Executor

    Permalink
  31. def predict(data: DataIter, numBatch: Int = 1): Array[NDArray]

    Permalink

    Run the prediction, always only use one device.

    Run the prediction, always only use one device.

    data

    eval data

    numBatch

    the number of batch to run. Go though all batches if set -1

    returns

    The predicted value of the output. Note the network may have multiple outputs, thus it return an array of NDArray

  32. def save(prefix: String, epoch: Int = this.numEpoch): Unit

    Permalink

    Checkpoint the model checkpoint into file.

    Checkpoint the model checkpoint into file. You can also use pickle to do the job if you only work on python. The advantage of load/save is the file is language agnostic. This means the file saved using save can be loaded by other language binding of mxnet. You also get the benefit being able to directly load/save from cloud storage(S3, HDFS)

    prefix

    Prefix of model name.

    Note

    - prefix-symbol.json will be saved for symbol. - prefix-epoch.params will be saved for parameters.

    See also

    FeedForward.load : the method to load the model back.

  33. def serialize(): Array[Byte]

    Permalink

    Serialize the model to Java byte array

    Serialize the model to Java byte array

    returns

    serialized model bytes

  34. def setMonitor(m: Monitor): Unit

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

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

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

    Permalink
  38. final def wait(): Unit

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

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

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

Inherited from AnyRef

Inherited from Any

Ungrouped