All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Modules Pages
BlockAffineComponent Class Reference

This class implements an affine transform using a block diagonal matrix e.g., one whose weight matrix is all zeros except for blocks on the diagonal. More...

#include <nnet-simple-component.h>

Inheritance diagram for BlockAffineComponent:
Collaboration diagram for BlockAffineComponent:

Public Member Functions

virtual int32 InputDim () const
 Returns input-dimension of this component. More...
 
virtual int32 OutputDim () const
 Returns output-dimension of this component. More...
 
virtual std::string Info () const
 Returns some text-form information about this component, for diagnostics. More...
 
virtual void InitFromConfig (ConfigLine *cfl)
 Initialize, from a ConfigLine object. More...
 
 BlockAffineComponent ()
 
virtual std::string Type () const
 Returns a string such as "SigmoidComponent", describing the type of the object. More...
 
virtual int32 Properties () const
 Return bitmask of the component's properties. More...
 
virtual void * Propagate (const ComponentPrecomputedIndexes *indexes, const CuMatrixBase< BaseFloat > &in, CuMatrixBase< BaseFloat > *out) const
 Propagate function. More...
 
virtual void Backprop (const std::string &debug_info, const ComponentPrecomputedIndexes *indexes, const CuMatrixBase< BaseFloat > &in_value, const CuMatrixBase< BaseFloat > &, const CuMatrixBase< BaseFloat > &out_deriv, void *memo, Component *to_update, CuMatrixBase< BaseFloat > *in_deriv) const
 Backprop function; depending on which of the arguments 'to_update' and 'in_deriv' are non-NULL, this can compute input-data derivatives and/or perform model update. More...
 
virtual void Read (std::istream &is, bool binary)
 Read function (used after we know the type of the Component); accepts input that is missing the token that describes the component type, in case it has already been consumed. More...
 
virtual void Write (std::ostream &os, bool binary) const
 Write component to stream. More...
 
virtual ComponentCopy () const
 Copies component (deep copy). More...
 
virtual void Scale (BaseFloat scale)
 This virtual function when called on – an UpdatableComponent scales the parameters by "scale" when called by an UpdatableComponent. More...
 
virtual void Add (BaseFloat alpha, const Component &other)
 This virtual function when called by – an UpdatableComponent adds the parameters of another updatable component, times some constant, to the current parameters. More...
 
virtual void PerturbParams (BaseFloat stddev)
 This function is to be used in testing. More...
 
virtual BaseFloat DotProduct (const UpdatableComponent &other) const
 Computes dot-product between parameters of two instances of a Component. More...
 
virtual int32 NumParameters () const
 The following new virtual function returns the total dimension of the parameters in this class. More...
 
virtual void Vectorize (VectorBase< BaseFloat > *params) const
 Turns the parameters into vector form. More...
 
virtual void UnVectorize (const VectorBase< BaseFloat > &params)
 Converts the parameters from vector form. More...
 
 BlockAffineComponent (const BlockAffineComponent &other)
 
 BlockAffineComponent (const RepeatedAffineComponent &rac)
 
- Public Member Functions inherited from UpdatableComponent
 UpdatableComponent (const UpdatableComponent &other)
 
 UpdatableComponent ()
 
virtual ~UpdatableComponent ()
 
virtual void SetUnderlyingLearningRate (BaseFloat lrate)
 Sets the learning rate of gradient descent- gets multiplied by learning_rate_factor_. More...
 
virtual void SetActualLearningRate (BaseFloat lrate)
 Sets the learning rate directly, bypassing learning_rate_factor_. More...
 
virtual void SetAsGradient ()
 Sets is_gradient_ to true and sets learning_rate_ to 1, ignoring learning_rate_factor_. More...
 
virtual BaseFloat LearningRateFactor ()
 
virtual void SetLearningRateFactor (BaseFloat lrate_factor)
 
void SetUpdatableConfigs (const UpdatableComponent &other)
 
virtual void FreezeNaturalGradient (bool freeze)
 freezes/unfreezes NaturalGradient updates, if applicable (to be overriden by components that use Natural Gradient). More...
 
BaseFloat LearningRate () const
 Gets the learning rate to be used in gradient descent. More...
 
BaseFloat MaxChange () const
 Returns the per-component max-change value, which is interpreted as the maximum change (in l2 norm) in parameters that is allowed per minibatch for this component. More...
 
void SetMaxChange (BaseFloat max_change)
 
BaseFloat L2Regularization () const
 Returns the l2 regularization constant, which may be set in any updatable component (usually from the config file). More...
 
void SetL2Regularization (BaseFloat a)
 
- Public Member Functions inherited from Component
virtual void StoreStats (const CuMatrixBase< BaseFloat > &in_value, const CuMatrixBase< BaseFloat > &out_value, void *memo)
 This function may store stats on average activation values, and for some component types, the average value of the derivative of the nonlinearity. More...
 
virtual void ZeroStats ()
 Components that provide an implementation of StoreStats should also provide an implementation of ZeroStats(), to set those stats to zero. More...
 
virtual void GetInputIndexes (const MiscComputationInfo &misc_info, const Index &output_index, std::vector< Index > *desired_indexes) const
 This function only does something interesting for non-simple Components. More...
 
virtual bool IsComputable (const MiscComputationInfo &misc_info, const Index &output_index, const IndexSet &input_index_set, std::vector< Index > *used_inputs) const
 This function only does something interesting for non-simple Components, and it exists to make it possible to manage optionally-required inputs. More...
 
virtual void ReorderIndexes (std::vector< Index > *input_indexes, std::vector< Index > *output_indexes) const
 This function only does something interesting for non-simple Components. More...
 
virtual
ComponentPrecomputedIndexes
PrecomputeIndexes (const MiscComputationInfo &misc_info, const std::vector< Index > &input_indexes, const std::vector< Index > &output_indexes, bool need_backprop) const
 This function must return NULL for simple Components. More...
 
virtual void DeleteMemo (void *memo) const
 This virtual function only needs to be overwritten by Components that return a non-NULL memo from their Propagate() function. More...
 
 Component ()
 
virtual ~Component ()
 

Protected Attributes

CuMatrix< BaseFloatlinear_params_
 
CuVector< BaseFloatbias_params_
 
int32 num_blocks_
 
- Protected Attributes inherited from UpdatableComponent
BaseFloat learning_rate_
 learning rate (typically 0.0..0.01) More...
 
BaseFloat learning_rate_factor_
 learning rate factor (normally 1.0, but can be set to another < value so that when < you call SetLearningRate(), that value will be scaled by this factor. More...
 
BaseFloat l2_regularize_
 L2 regularization constant. More...
 
bool is_gradient_
 True if this component is to be treated as a gradient rather than as parameters. More...
 
BaseFloat max_change_
 configuration value for imposing max-change More...
 

Private Member Functions

void Init (int32 input_dim, int32 output_dim, int32 num_blocks, BaseFloat param_stddev, BaseFloat bias_mean, BaseFloat bias_stddev)
 
const BlockAffineComponentoperator= (const BlockAffineComponent &other)
 

Additional Inherited Members

- Static Public Member Functions inherited from Component
static ComponentReadNew (std::istream &is, bool binary)
 Read component from stream (works out its type). Dies on error. More...
 
static ComponentNewComponentOfType (const std::string &type)
 Returns a new Component of the given type e.g. More...
 
- Protected Member Functions inherited from UpdatableComponent
void InitLearningRatesFromConfig (ConfigLine *cfl)
 
std::string ReadUpdatableCommon (std::istream &is, bool binary)
 
void WriteUpdatableCommon (std::ostream &is, bool binary) const
 

Detailed Description

This class implements an affine transform using a block diagonal matrix e.g., one whose weight matrix is all zeros except for blocks on the diagonal.

All these blocks have the same dimensions. input-dim: num cols of block diagonal matrix. output-dim: num rows of block diagonal matrix. num-blocks: number of blocks in diagonal of the matrix. num-blocks must divide both input-dim and output-dim

Definition at line 483 of file nnet-simple-component.h.

Constructor & Destructor Documentation

Definition at line 491 of file nnet-simple-component.h.

Referenced by BlockAffineComponent::Copy().

491 { }
BlockAffineComponent ( const BlockAffineComponent other)
explicit

Definition at line 1564 of file nnet-simple-component.cc.

1564  :
1565  UpdatableComponent(other),
1566  linear_params_(other.linear_params_),
1567  bias_params_(other.bias_params_),
1568  num_blocks_(other.num_blocks_) {}
BlockAffineComponent ( const RepeatedAffineComponent rac)
explicit

Definition at line 1570 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, RepeatedAffineComponent::bias_params_, CuMatrixBase< Real >::CopyFromMat(), CuVectorBase< Real >::CopyFromVec(), BlockAffineComponent::linear_params_, RepeatedAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, CuMatrixBase< Real >::NumRows(), CuVectorBase< Real >::Range(), and CuMatrixBase< Real >::RowRange().

1570  :
1571  UpdatableComponent(rac),
1572  linear_params_(rac.num_repeats_ * rac.linear_params_.NumRows(),
1573  rac.linear_params_.NumCols(), kUndefined),
1574  bias_params_(rac.num_repeats_ * rac.linear_params_.NumRows(), kUndefined),
1575  num_blocks_(rac.num_repeats_) {
1576  // copy rac's linear_params_ and bias_params_ to this.
1577  int32 num_rows_in_block = rac.linear_params_.NumRows();
1578  for(int32 block_counter = 0; block_counter < num_blocks_; block_counter++) {
1579  int32 row_offset = block_counter * num_rows_in_block;
1580  CuSubMatrix<BaseFloat> block = this->linear_params_.RowRange(row_offset,
1581  num_rows_in_block);
1582  block.CopyFromMat(rac.linear_params_);
1583  CuSubVector<BaseFloat> block_bias = this->bias_params_.Range(row_offset,
1584  num_rows_in_block);
1585  block_bias.CopyFromVec(rac.bias_params_);
1586  }
1587 }
CuSubMatrix< Real > RowRange(const MatrixIndexT row_offset, const MatrixIndexT num_rows) const
Definition: cu-matrix.h:603
CuSubVector< Real > Range(const MatrixIndexT o, const MatrixIndexT l)
Definition: cu-vector.h:145

Member Function Documentation

void Add ( BaseFloat  alpha,
const Component other 
)
virtual

This virtual function when called by – an UpdatableComponent adds the parameters of another updatable component, times some constant, to the current parameters.

– a NonlinearComponent (or another component that stores stats, like BatchNormComponent)– it relates to adding stats. Otherwise it will normally do nothing.

Reimplemented from Component.

Definition at line 1774 of file nnet-simple-component.cc.

References CuMatrixBase< Real >::AddMat(), CuVectorBase< Real >::AddVec(), BlockAffineComponent::bias_params_, KALDI_ASSERT, and BlockAffineComponent::linear_params_.

1774  {
1775  const BlockAffineComponent *other =
1776  dynamic_cast<const BlockAffineComponent *>(&other_in);
1777  KALDI_ASSERT(other != NULL);
1778  linear_params_.AddMat(alpha, other->linear_params_);
1779  bias_params_.AddVec(alpha, other->bias_params_);
1780 }
void AddMat(Real alpha, const CuMatrixBase< Real > &A, MatrixTransposeType trans=kNoTrans)
*this += alpha * A
Definition: cu-matrix.cc:941
#define KALDI_ASSERT(cond)
Definition: kaldi-error.h:169
void AddVec(Real alpha, const CuVectorBase< Real > &vec, Real beta=1.0)
Definition: cu-vector.cc:1182
void Backprop ( const std::string &  debug_info,
const ComponentPrecomputedIndexes indexes,
const CuMatrixBase< BaseFloat > &  in_value,
const CuMatrixBase< BaseFloat > &  out_value,
const CuMatrixBase< BaseFloat > &  out_deriv,
void *  memo,
Component to_update,
CuMatrixBase< BaseFloat > *  in_deriv 
) const
virtual

Backprop function; depending on which of the arguments 'to_update' and 'in_deriv' are non-NULL, this can compute input-data derivatives and/or perform model update.

Parameters
[in]debug_infoThe component name, to be printed out in any warning messages.
[in]indexesA pointer to some information output by this class's PrecomputeIndexes function (will be NULL for simple components, i.e. those that don't do things like splicing).
[in]in_valueThe matrix that was given as input to the Propagate function. Will be ignored (and may be empty) if Properties()&kBackpropNeedsInput == 0.
[in]out_valueThe matrix that was output from the Propagate function. Will be ignored (and may be empty) if Properties()&kBackpropNeedsOutput == 0
[in]out_derivThe derivative at the output of this component.
[in]memoThis will normally be NULL, but for component types that set the flag kUsesMemo, this will be the return value of the Propagate() function that corresponds to this Backprop() function. Ownership of any pointers is not transferred to the Backprop function; DeleteMemo() will be called to delete it.
[out]to_updateIf model update is desired, the Component to be updated, else NULL. Does not have to be identical to this. If supplied, you can assume that to_update->Properties() & kUpdatableComponent is nonzero.
[out]in_derivThe derivative at the input of this component, if needed (else NULL). If Properties()&kBackpropInPlace, may be the same matrix as out_deriv. If Properties()&kBackpropAdds, this is added to by the Backprop routine, else it is set. The component code chooses which mode to work in, based on convenience.

Implements Component.

Definition at line 1678 of file nnet-simple-component.cc.

References CuVectorBase< Real >::AddRowSumMat(), BlockAffineComponent::bias_params_, CuMatrixBase< Real >::ColRange(), kaldi::DeletePointers(), kaldi::kNoTrans, kaldi::kTrans, UpdatableComponent::learning_rate_, BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, CuMatrixBase< Real >::NumCols(), CuMatrixBase< Real >::NumRows(), and CuMatrixBase< Real >::RowRange().

1685  {
1686  BlockAffineComponent *to_update = dynamic_cast<BlockAffineComponent*>(to_update_in);
1687 
1688  const int32 num_rows_in_block = linear_params_.NumRows() / num_blocks_;
1689  const int32 num_cols_in_block = linear_params_.NumCols();
1690 
1691  // Propagate the derivative back to the input.
1692  // add with coefficient 1.0 since property kBackpropAdds is true.
1693  // If we wanted to add with coefficient 0.0 we'd need to zero the
1694  // in_deriv, in case of infinities.
1695  if (in_deriv) {
1696  std::vector<CuSubMatrix<BaseFloat> *> in_deriv_batch, out_deriv_batch, linear_params_batch;
1697 
1698  for(int block_counter = 0; block_counter < num_blocks_; block_counter++) {
1699  CuSubMatrix<BaseFloat> *in_deriv_block =
1700  new CuSubMatrix<BaseFloat>(in_deriv->ColRange(block_counter * num_cols_in_block,
1701  num_cols_in_block));
1702  in_deriv_batch.push_back(in_deriv_block);
1703 
1704  CuSubMatrix<BaseFloat> *out_deriv_block =
1705  new CuSubMatrix<BaseFloat>(out_deriv.ColRange(block_counter * num_rows_in_block,
1706  num_rows_in_block));
1707  out_deriv_batch.push_back(out_deriv_block);
1708 
1709  CuSubMatrix<BaseFloat> *linear_params_block =
1710  new CuSubMatrix<BaseFloat>(linear_params_.RowRange(block_counter * num_rows_in_block,
1711  num_rows_in_block));
1712  linear_params_batch.push_back(linear_params_block);
1713  }
1714 
1715  AddMatMatBatched<BaseFloat>(1.0, in_deriv_batch, out_deriv_batch, kNoTrans,
1716  linear_params_batch, kNoTrans, 1.0);
1717 
1718  DeletePointers(&in_deriv_batch);
1719  DeletePointers(&out_deriv_batch);
1720  DeletePointers(&linear_params_batch);
1721  }
1722 
1723  if (to_update != NULL) {
1724 
1725  { // linear params update
1726 
1727  std::vector<CuSubMatrix<BaseFloat> *> in_value_batch,
1728  out_deriv_batch, linear_params_batch;
1729 
1730  for (int block_counter = 0; block_counter < num_blocks_; block_counter++) {
1731  CuSubMatrix<BaseFloat> *in_value_block =
1732  new CuSubMatrix<BaseFloat>(in_value.ColRange(block_counter * num_cols_in_block,
1733  num_cols_in_block));
1734  in_value_batch.push_back(in_value_block);
1735 
1736  CuSubMatrix<BaseFloat> *out_deriv_block =
1737  new CuSubMatrix<BaseFloat>(out_deriv.ColRange(block_counter * num_rows_in_block,
1738  num_rows_in_block));
1739  out_deriv_batch.push_back(out_deriv_block);
1740 
1741  CuSubMatrix<BaseFloat> *linear_params_block =
1742  new CuSubMatrix<BaseFloat>(to_update->linear_params_.RowRange(block_counter * num_rows_in_block,
1743  num_rows_in_block));
1744  linear_params_batch.push_back(linear_params_block);
1745  }
1746 
1747  AddMatMatBatched<BaseFloat>(to_update->learning_rate_,
1748  linear_params_batch,
1749  out_deriv_batch, kTrans,
1750  in_value_batch, kNoTrans, 1.0);
1751 
1752  DeletePointers(&in_value_batch);
1753  DeletePointers(&out_deriv_batch);
1754  DeletePointers(&linear_params_batch);
1755  } // end linear params update
1756 
1757  { // bias update
1758  to_update->bias_params_.AddRowSumMat(to_update->learning_rate_,
1759  out_deriv, 1.0);
1760  } // end bias update
1761  }
1762 }
MatrixIndexT NumCols() const
Definition: cu-matrix.h:215
CuSubMatrix< Real > RowRange(const MatrixIndexT row_offset, const MatrixIndexT num_rows) const
Definition: cu-matrix.h:603
CuSubMatrix< Real > ColRange(const MatrixIndexT col_offset, const MatrixIndexT num_cols) const
Definition: cu-matrix.h:608
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
void DeletePointers(std::vector< A * > *v)
Deletes any non-NULL pointers in the vector v, and sets the corresponding entries of v to NULL...
Definition: stl-utils.h:186
Component * Copy ( ) const
virtual

Copies component (deep copy).

Implements Component.

Definition at line 1589 of file nnet-simple-component.cc.

References BlockAffineComponent::BlockAffineComponent().

1589  {
1590  BlockAffineComponent *ans = new BlockAffineComponent(*this);
1591  return ans;
1592 }
BaseFloat DotProduct ( const UpdatableComponent other) const
virtual

Computes dot-product between parameters of two instances of a Component.

Can be used for computing parameter-norm of an UpdatableComponent.

Implements UpdatableComponent.

Definition at line 1792 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, kaldi::kTrans, BlockAffineComponent::linear_params_, kaldi::TraceMatMat(), and kaldi::VecVec().

1792  {
1793  const BlockAffineComponent *other =
1794  dynamic_cast<const BlockAffineComponent*>(&other_in);
1795  return TraceMatMat(linear_params_, other->linear_params_, kTrans) +
1796  VecVec(bias_params_, other->bias_params_);
1797 }
Real TraceMatMat(const MatrixBase< Real > &A, const MatrixBase< Real > &B, MatrixTransposeType trans=kNoTrans)
We need to declare this here as it will be a friend function.
Real VecVec(const VectorBase< Real > &a, const VectorBase< Real > &b)
Returns dot product between v1 and v2.
Definition: kaldi-vector.cc:37
std::string Info ( ) const
virtual

Returns some text-form information about this component, for diagnostics.

Starts with the type of the component. E.g. "SigmoidComponent dim=900", although most components will have much more info.

Reimplemented from UpdatableComponent.

Definition at line 1594 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, UpdatableComponent::Info(), BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, and kaldi::nnet3::PrintParameterStats().

1594  {
1595  std::ostringstream stream;
1596  stream << UpdatableComponent::Info()
1597  << ", num-blocks=" << num_blocks_;
1598  PrintParameterStats(stream, "linear-params", linear_params_);
1599  PrintParameterStats(stream, "bias", bias_params_, true);
1600  return stream.str();
1601 }
virtual std::string Info() const
Returns some text-form information about this component, for diagnostics.
void PrintParameterStats(std::ostringstream &os, const std::string &name, const CuVectorBase< BaseFloat > &params, bool include_mean)
Print to 'os' some information about the mean and standard deviation of some parameters, used in Info() functions in nnet-simple-component.cc.
Definition: nnet-parse.cc:530
void Init ( int32  input_dim,
int32  output_dim,
int32  num_blocks,
BaseFloat  param_stddev,
BaseFloat  bias_mean,
BaseFloat  bias_stddev 
)
private

Definition at line 1603 of file nnet-simple-component.cc.

References CuVectorBase< Real >::Add(), BlockAffineComponent::bias_params_, KALDI_ASSERT, BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, CuVector< Real >::Resize(), CuMatrix< Real >::Resize(), CuVectorBase< Real >::Scale(), CuMatrixBase< Real >::Scale(), CuVectorBase< Real >::SetRandn(), and CuMatrixBase< Real >::SetRandn().

Referenced by BlockAffineComponent::InitFromConfig().

1606  {
1607  KALDI_ASSERT(input_dim > 0 && output_dim > 0 && num_blocks >= 1);
1608  KALDI_ASSERT(output_dim % num_blocks == 0 && input_dim % num_blocks == 0);
1609  const int32 num_columns_per_block = input_dim / num_blocks;
1610  linear_params_.Resize(output_dim, num_columns_per_block);
1611  bias_params_.Resize(output_dim);
1612  KALDI_ASSERT(param_stddev >= 0.0 && bias_stddev >= 0.0);
1614  linear_params_.Scale(param_stddev);
1616  bias_params_.Scale(bias_stddev);
1617  bias_params_.Add(bias_mean);
1618  num_blocks_ = num_blocks;
1619 }
void Scale(Real value)
Definition: cu-vector.cc:1161
void Scale(Real value)
Definition: cu-matrix.cc:610
void Resize(MatrixIndexT dim, MatrixResizeType t=kSetZero)
Allocate the memory.
Definition: cu-vector.cc:941
void Resize(MatrixIndexT rows, MatrixIndexT cols, MatrixResizeType resize_type=kSetZero, MatrixStrideType stride_type=kDefaultStride)
Allocate the memory.
Definition: cu-matrix.cc:49
void Add(Real value)
Definition: cu-vector.cc:1102
#define KALDI_ASSERT(cond)
Definition: kaldi-error.h:169
void InitFromConfig ( ConfigLine cfl)
virtual

Initialize, from a ConfigLine object.

Parameters
[in]cflA ConfigLine containing any parameters that are needed for initialization. For example: "dim=100 param-stddev=0.1"

Implements Component.

Definition at line 1621 of file nnet-simple-component.cc.

References ConfigLine::GetValue(), ConfigLine::HasUnusedValues(), BlockAffineComponent::Init(), UpdatableComponent::InitLearningRatesFromConfig(), KALDI_ERR, BlockAffineComponent::Type(), and ConfigLine::WholeLine().

1621  {
1622  int32 input_dim = -1, output_dim = -1, num_blocks = -1;
1623  if(!cfl->GetValue("input-dim", &input_dim) ||
1624  !cfl->GetValue("output-dim", &output_dim) ||
1625  !cfl->GetValue("num-blocks", &num_blocks))
1626  KALDI_ERR << "Invalid initializer for layer of type "
1627  << Type() << ": \"" << cfl->WholeLine() << "\"";
1629  BaseFloat param_stddev = 1.0 / std::sqrt(input_dim / num_blocks),
1630  bias_mean = 0.0, bias_stddev = 1.0;
1631  cfl->GetValue("param-stddev", &param_stddev);
1632  cfl->GetValue("bias-stddev", &bias_stddev);
1633  cfl->GetValue("bias-mean", &bias_mean);
1634 
1635  if (cfl->HasUnusedValues())
1636  KALDI_ERR << "Invalid initializer for layer of type "
1637  << Type() << ": \"" << cfl->WholeLine() << "\"";
1638 
1639  Init(input_dim, output_dim, num_blocks,
1640  param_stddev, bias_mean, bias_stddev);
1641 }
void InitLearningRatesFromConfig(ConfigLine *cfl)
float BaseFloat
Definition: kaldi-types.h:29
void Init(int32 input_dim, int32 output_dim, int32 num_blocks, BaseFloat param_stddev, BaseFloat bias_mean, BaseFloat bias_stddev)
#define KALDI_ERR
Definition: kaldi-error.h:127
virtual std::string Type() const
Returns a string such as "SigmoidComponent", describing the type of the object.
virtual int32 InputDim ( ) const
inlinevirtual

Returns input-dimension of this component.

Implements Component.

Definition at line 485 of file nnet-simple-component.h.

References BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, and CuMatrixBase< Real >::NumCols().

int32 NumParameters ( ) const
virtual

The following new virtual function returns the total dimension of the parameters in this class.

Reimplemented from UpdatableComponent.

Definition at line 1827 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, CuVectorBase< Real >::Dim(), BlockAffineComponent::linear_params_, CuMatrixBase< Real >::NumCols(), and CuMatrixBase< Real >::NumRows().

Referenced by BlockAffineComponent::UnVectorize(), and BlockAffineComponent::Vectorize().

1827  {
1829 }
MatrixIndexT NumCols() const
Definition: cu-matrix.h:215
MatrixIndexT Dim() const
Dimensions.
Definition: cu-vector.h:68
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
const BlockAffineComponent& operator= ( const BlockAffineComponent other)
private
virtual int32 OutputDim ( ) const
inlinevirtual

Returns output-dimension of this component.

Implements Component.

Definition at line 486 of file nnet-simple-component.h.

References BlockAffineComponent::linear_params_, and CuMatrixBase< Real >::NumRows().

486 { return linear_params_.NumRows(); }
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
void PerturbParams ( BaseFloat  stddev)
virtual

This function is to be used in testing.

It adds unit noise times "stddev" to the parameters of the component.

Implements UpdatableComponent.

Definition at line 1782 of file nnet-simple-component.cc.

References CuMatrixBase< Real >::AddMat(), CuVectorBase< Real >::AddVec(), BlockAffineComponent::bias_params_, BlockAffineComponent::linear_params_, CuVectorBase< Real >::SetRandn(), and CuMatrixBase< Real >::SetRandn().

1782  {
1783  CuMatrix<BaseFloat> temp_linear_params(linear_params_);
1784  temp_linear_params.SetRandn();
1785  linear_params_.AddMat(stddev, temp_linear_params);
1786 
1787  CuVector<BaseFloat> temp_bias_params(bias_params_);
1788  temp_bias_params.SetRandn();
1789  bias_params_.AddVec(stddev, temp_bias_params);
1790 }
void AddMat(Real alpha, const CuMatrixBase< Real > &A, MatrixTransposeType trans=kNoTrans)
*this += alpha * A
Definition: cu-matrix.cc:941
void AddVec(Real alpha, const CuVectorBase< Real > &vec, Real beta=1.0)
Definition: cu-vector.cc:1182
void * Propagate ( const ComponentPrecomputedIndexes indexes,
const CuMatrixBase< BaseFloat > &  in,
CuMatrixBase< BaseFloat > *  out 
) const
virtual

Propagate function.

Parameters
[in]indexesA pointer to some information output by this class's PrecomputeIndexes function (will be NULL for simple components, i.e. those that don't do things like splicing).
[in]inThe input to this component. Num-columns == InputDim().
[out]outThe output of this component. Num-columns == OutputDim(). Note: output of this component will be added to the initial value of "out" if Properties()&kPropagateAdds != 0; otherwise the output will be set and the initial value ignored. Each Component chooses whether it is more convenient implementation-wise to add or set, and the calling code has to deal with it.
Returns
Normally returns NULL, but may return a non-NULL value for components which have the flag kUsesMemo set. This value will be passed into the corresponding Backprop routine.

Implements Component.

Definition at line 1643 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, CuMatrixBase< Real >::ColRange(), CuMatrixBase< Real >::CopyRowsFromVec(), kaldi::DeletePointers(), kaldi::kNoTrans, kaldi::kTrans, BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, CuMatrixBase< Real >::NumCols(), CuMatrixBase< Real >::NumRows(), and CuMatrixBase< Real >::RowRange().

1645  {
1647  // block_dimension is both the number of columns, and the number of rows,
1648  // of a block.
1649  int32 num_rows_in_block = linear_params_.NumRows() / num_blocks_;
1650  int32 num_cols_in_block = linear_params_.NumCols();
1651  std::vector<CuSubMatrix<BaseFloat> *> in_batch, out_batch,
1652  linear_params_batch;
1653  for(int block_counter = 0; block_counter < num_blocks_; block_counter++) {
1654  CuSubMatrix<BaseFloat> *in_block =
1655  new CuSubMatrix<BaseFloat>(in.ColRange(block_counter * num_cols_in_block,
1656  num_cols_in_block));
1657  in_batch.push_back(in_block);
1658 
1659  CuSubMatrix<BaseFloat> *out_block =
1660  new CuSubMatrix<BaseFloat>(out->ColRange(block_counter * num_rows_in_block,
1661  num_rows_in_block));
1662  out_batch.push_back(out_block);
1663 
1664  CuSubMatrix<BaseFloat> *linear_params_block =
1665  new CuSubMatrix<BaseFloat>(linear_params_.RowRange(block_counter * num_rows_in_block,
1666  num_rows_in_block));
1667  linear_params_batch.push_back(linear_params_block);
1668  }
1669  AddMatMatBatched<BaseFloat>(1.0, out_batch, in_batch, kNoTrans,
1670  linear_params_batch, kTrans, 1.0);
1671 
1672  DeletePointers(&in_batch);
1673  DeletePointers(&out_batch);
1674  DeletePointers(&linear_params_batch);
1675  return NULL;
1676 }
void CopyRowsFromVec(const CuVectorBase< Real > &v)
This function has two modes of operation.
Definition: cu-matrix.cc:2282
MatrixIndexT NumCols() const
Definition: cu-matrix.h:215
CuSubMatrix< Real > RowRange(const MatrixIndexT row_offset, const MatrixIndexT num_rows) const
Definition: cu-matrix.h:603
CuSubMatrix< Real > ColRange(const MatrixIndexT col_offset, const MatrixIndexT num_cols) const
Definition: cu-matrix.h:608
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
void DeletePointers(std::vector< A * > *v)
Deletes any non-NULL pointers in the vector v, and sets the corresponding entries of v to NULL...
Definition: stl-utils.h:186
virtual int32 Properties ( ) const
inlinevirtual
void Read ( std::istream &  is,
bool  binary 
)
virtual

Read function (used after we know the type of the Component); accepts input that is missing the token that describes the component type, in case it has already been consumed.

Implements Component.

Definition at line 1799 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, kaldi::nnet3::ExpectToken(), UpdatableComponent::is_gradient_, BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, kaldi::PeekToken(), CuVector< Real >::Read(), CuMatrix< Real >::Read(), kaldi::ReadBasicType(), and UpdatableComponent::ReadUpdatableCommon().

1799  {
1800  ReadUpdatableCommon(is, binary); // read opening tag and learning rate.
1801  ExpectToken(is, binary, "<NumBlocks>");
1802  ReadBasicType(is, binary, &num_blocks_);
1803  ExpectToken(is, binary, "<LinearParams>");
1804  linear_params_.Read(is, binary);
1805  ExpectToken(is, binary, "<BiasParams>");
1806  bias_params_.Read(is, binary);
1807  if (PeekToken(is, binary) == 'I') {
1808  // for back compatibility; we don't write this here any
1809  // more as it's written and read in Write/ReadUpdatableCommon
1810  ExpectToken(is, binary, "<IsGradient>");
1811  ReadBasicType(is, binary, &is_gradient_);
1812  }
1813  ExpectToken(is, binary, "</BlockAffineComponent>");
1814 }
void ReadBasicType(std::istream &is, bool binary, T *t)
ReadBasicType is the name of the read function for bool, integer types, and floating-point types...
Definition: io-funcs-inl.h:55
void Read(std::istream &is, bool binary)
I/O.
Definition: cu-vector.cc:911
static void ExpectToken(const std::string &token, const std::string &what_we_are_parsing, const std::string **next_token)
std::string ReadUpdatableCommon(std::istream &is, bool binary)
int PeekToken(std::istream &is, bool binary)
PeekToken will return the first character of the next token, or -1 if end of file.
Definition: io-funcs.cc:170
void Read(std::istream &is, bool binary)
I/O functions.
Definition: cu-matrix.cc:461
bool is_gradient_
True if this component is to be treated as a gradient rather than as parameters.
void Scale ( BaseFloat  scale)
virtual

This virtual function when called on – an UpdatableComponent scales the parameters by "scale" when called by an UpdatableComponent.

– a Nonlinear component (or another component that stores stats, like BatchNormComponent)– it relates to scaling activation stats, not parameters. Otherwise it will normally do nothing.

Reimplemented from Component.

Definition at line 1764 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, BlockAffineComponent::linear_params_, CuVectorBase< Real >::Scale(), CuMatrixBase< Real >::Scale(), CuVectorBase< Real >::SetZero(), and CuMatrixBase< Real >::SetZero().

1764  {
1765  if (scale == 0.0) {
1768  } else {
1769  linear_params_.Scale(scale);
1770  bias_params_.Scale(scale);
1771  }
1772 }
void Scale(Real value)
Definition: cu-vector.cc:1161
void Scale(Real value)
Definition: cu-matrix.cc:610
void SetZero()
Math operations, some calling kernels.
Definition: cu-matrix.cc:476
void SetZero()
Math operations.
Definition: cu-vector.cc:1044
virtual std::string Type ( ) const
inlinevirtual

Returns a string such as "SigmoidComponent", describing the type of the object.

Implements Component.

Definition at line 492 of file nnet-simple-component.h.

Referenced by BlockAffineComponent::InitFromConfig().

492 { return "BlockAffineComponent"; }
void UnVectorize ( const VectorBase< BaseFloat > &  params)
virtual

Converts the parameters from vector form.

Reimplemented from UpdatableComponent.

Definition at line 1839 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, CuVectorBase< Real >::CopyFromVec(), CuMatrixBase< Real >::CopyRowsFromVec(), VectorBase< Real >::Dim(), CuVectorBase< Real >::Dim(), KALDI_ASSERT, BlockAffineComponent::linear_params_, CuMatrixBase< Real >::NumCols(), BlockAffineComponent::NumParameters(), CuMatrixBase< Real >::NumRows(), and VectorBase< Real >::Range().

1839  {
1840  KALDI_ASSERT(params.Dim() == this->NumParameters());
1841  int32 num_linear_params = linear_params_.NumCols() * linear_params_.NumRows();
1842  int32 num_bias_params = bias_params_.Dim();
1843  linear_params_.CopyRowsFromVec(params.Range(0, num_linear_params));
1844  bias_params_.CopyFromVec(params.Range(num_linear_params, num_bias_params));
1845 }
void CopyRowsFromVec(const CuVectorBase< Real > &v)
This function has two modes of operation.
Definition: cu-matrix.cc:2282
MatrixIndexT NumCols() const
Definition: cu-matrix.h:215
MatrixIndexT Dim() const
Dimensions.
Definition: cu-vector.h:68
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
void CopyFromVec(const CuVectorBase< Real > &src)
Copy functions; these will crash if the dimension do not match.
Definition: cu-vector.cc:1026
virtual int32 NumParameters() const
The following new virtual function returns the total dimension of the parameters in this class...
#define KALDI_ASSERT(cond)
Definition: kaldi-error.h:169
MatrixIndexT Dim() const
Returns the dimension of the vector.
Definition: kaldi-vector.h:63
SubVector< Real > Range(const MatrixIndexT o, const MatrixIndexT l)
Returns a sub-vector of a vector (a range of elements).
Definition: kaldi-vector.h:93
void Vectorize ( VectorBase< BaseFloat > *  params) const
virtual

Turns the parameters into vector form.

We put the vector form on the CPU, because in the kinds of situations where we do this, we'll tend to use too much memory for the GPU.

Reimplemented from UpdatableComponent.

Definition at line 1831 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, VectorBase< Real >::Dim(), CuVectorBase< Real >::Dim(), KALDI_ASSERT, BlockAffineComponent::linear_params_, CuMatrixBase< Real >::NumCols(), BlockAffineComponent::NumParameters(), CuMatrixBase< Real >::NumRows(), and VectorBase< Real >::Range().

1831  {
1832  KALDI_ASSERT(params->Dim() == this->NumParameters());
1833  int32 num_linear_params = linear_params_.NumCols() * linear_params_.NumRows();
1834  int32 num_bias_params = bias_params_.Dim();
1835  params->Range(0, num_linear_params).CopyRowsFromMat(linear_params_);
1836  params->Range(num_linear_params, num_bias_params).CopyFromVec(bias_params_);
1837 }
MatrixIndexT NumCols() const
Definition: cu-matrix.h:215
MatrixIndexT Dim() const
Dimensions.
Definition: cu-vector.h:68
MatrixIndexT NumRows() const
Dimensions.
Definition: cu-matrix.h:214
virtual int32 NumParameters() const
The following new virtual function returns the total dimension of the parameters in this class...
#define KALDI_ASSERT(cond)
Definition: kaldi-error.h:169
MatrixIndexT Dim() const
Returns the dimension of the vector.
Definition: kaldi-vector.h:63
SubVector< Real > Range(const MatrixIndexT o, const MatrixIndexT l)
Returns a sub-vector of a vector (a range of elements).
Definition: kaldi-vector.h:93
void Write ( std::ostream &  os,
bool  binary 
) const
virtual

Write component to stream.

Implements Component.

Definition at line 1816 of file nnet-simple-component.cc.

References BlockAffineComponent::bias_params_, BlockAffineComponent::linear_params_, BlockAffineComponent::num_blocks_, CuVector< Real >::Write(), CuMatrixBase< Real >::Write(), kaldi::WriteBasicType(), kaldi::WriteToken(), and UpdatableComponent::WriteUpdatableCommon().

1816  {
1817  WriteUpdatableCommon(os, binary); // Write opening tag and learning rate
1818  WriteToken(os, binary, "<NumBlocks>");
1819  WriteBasicType(os, binary, num_blocks_);
1820  WriteToken(os, binary, "<LinearParams>");
1821  linear_params_.Write(os, binary);
1822  WriteToken(os, binary, "<BiasParams>");
1823  bias_params_.Write(os, binary);
1824  WriteToken(os, binary, "</BlockAffineComponent>");
1825 }
void Write(std::ostream &is, bool binary) const
Definition: cu-vector.cc:921
void WriteUpdatableCommon(std::ostream &is, bool binary) const
void WriteToken(std::ostream &os, bool binary, const char *token)
The WriteToken functions are for writing nonempty sequences of non-space characters.
Definition: io-funcs.cc:134
void WriteBasicType(std::ostream &os, bool binary, T t)
WriteBasicType is the name of the write function for bool, integer types, and floating-point types...
Definition: io-funcs-inl.h:34
void Write(std::ostream &os, bool binary) const
Definition: cu-matrix.cc:469

Member Data Documentation


The documentation for this class was generated from the following files: