https://github.com/Microsoft/CNTK
Tip revision: 0cf886883fcddf780ef0093042e065c344c3eedd authored by thhoens on 23 November 2016, 00:56:56 UTC
Enabled 1D sparse convolution in Brainscript
Enabled 1D sparse convolution in Brainscript
Tip revision: 0cf8868
HTKDataDeserializer.cpp
//
// Copyright (c) Microsoft. All rights reserved.
// Licensed under the MIT license. See LICENSE.md file in the project root for full license information.
//
#include "stdafx.h"
#define __STDC_FORMAT_MACROS
#include <inttypes.h>
#include "HTKDataDeserializer.h"
#include "ConfigHelper.h"
#include "Basics.h"
#include "StringUtil.h"
// TODO: This will be removed when dependency on old code is eliminated.
// Currently this fixes the linking.
namespace msra { namespace asr {
std::unordered_map<std::wstring, unsigned int> htkfeatreader::parsedpath::archivePathStringMap;
std::vector<std::wstring> htkfeatreader::parsedpath::archivePathStringVector;
}}
namespace Microsoft { namespace MSR { namespace CNTK {
using namespace std;
HTKDataDeserializer::HTKDataDeserializer(
CorpusDescriptorPtr corpus,
const ConfigParameters& cfg,
bool primary)
: m_verbosity(0),
m_corpus(corpus),
m_primary(primary)
{
// TODO: This should be read in one place, potentially given by SGD.
m_frameMode = (ConfigValue)cfg("frameMode", "true");
m_verbosity = cfg(L"verbosity", 0);
argvector<ConfigValue> inputs = cfg("input");
if (inputs.size() != 1)
{
InvalidArgument("HTKDataDeserializer supports a single input stream only.");
}
ConfigParameters input = inputs.front();
auto inputName = input.GetMemberIds().front();
std::wstring precision = cfg(L"precision", L"float");
m_expandToPrimary = cfg(L"expandToUtterance", false);
if (m_expandToPrimary && m_primary)
{
InvalidArgument("Cannot expand utterances of the primary stream %ls, please change your configuration.", inputName.c_str());
}
ConfigParameters streamConfig = input(inputName);
ConfigHelper config(streamConfig);
auto context = config.GetContextWindow();
m_elementType = AreEqualIgnoreCase(precision, L"float") ? ElementType::tfloat : ElementType::tdouble;
m_dimension = config.GetFeatureDimension();
m_dimension = m_dimension * (1 + context.first + context.second);
InitializeChunkDescriptions(config);
InitializeStreams(inputName);
InitializeFeatureInformation();
InitializeAugmentationWindow(config);
}
HTKDataDeserializer::HTKDataDeserializer(
CorpusDescriptorPtr corpus,
const ConfigParameters& feature,
const wstring& featureName,
bool primary)
: m_corpus(corpus),
m_primary(primary)
{
// The frame mode is currently specified once per configuration,
// not in the configuration of a particular deserializer, but on a higher level in the configuration.
// Because of that we are using find method below.
m_frameMode = feature.Find("frameMode", "true");
ConfigHelper config(feature);
config.CheckFeatureType();
m_verbosity = feature(L"verbosity", 0);
auto context = config.GetContextWindow();
m_elementType = config.GetElementType();
m_dimension = config.GetFeatureDimension();
m_dimension = m_dimension * (1 + context.first + context.second);
m_expandToPrimary = feature(L"expandToUtterance", false);
if (m_expandToPrimary && m_primary)
{
InvalidArgument("Cannot expand utterances of the primary stream %ls, please change your configuration.", featureName.c_str());
}
InitializeChunkDescriptions(config);
InitializeStreams(featureName);
InitializeFeatureInformation();
InitializeAugmentationWindow(config);
}
void HTKDataDeserializer::InitializeAugmentationWindow(ConfigHelper& config)
{
m_augmentationWindow = config.GetContextWindow();
// If not given explicitly, we need to identify the required augmentation range from the expected dimension
// and the number of dimensions in the file.
if (m_augmentationWindow.first == 0 && m_augmentationWindow.second == 0)
{
m_augmentationWindow.first = m_augmentationWindow.second = msra::dbn::augmentationextent(m_ioFeatureDimension, m_dimension);
}
}
// Initializes chunks based on the configuration and utterance descriptions.
void HTKDataDeserializer::InitializeChunkDescriptions(ConfigHelper& config)
{
// Read utterance descriptions.
vector<wstring> paths = config.GetSequencePaths();
vector<UtteranceDescription> utterances;
utterances.reserve(paths.size());
auto& stringRegistry = m_corpus->GetStringRegistry();
size_t allUtterances = 0, allFrames = 0;
for (const auto& u : paths)
{
UtteranceDescription description(move(msra::asr::htkfeatreader::parsedpath(u)));
size_t numberOfFrames = description.GetNumberOfFrames();
if (m_expandToPrimary && numberOfFrames != 1)
{
RuntimeError("Expanded stream should only contain sequences of length 1, utterance '%s' has %d",
description.GetKey().c_str(),
(int)numberOfFrames);
}
// For logging, also account for utterances and frames that we skip
allUtterances++;
allFrames += numberOfFrames;
string key = description.GetKey();
if (!m_corpus->IsIncluded(key))
{
continue;
}
// No need to store key, releasing it.
description.ClearLogicalPath();
size_t id = stringRegistry[key];
description.SetId(id);
utterances.push_back(description);
m_totalNumberOfFrames += numberOfFrames;
}
// TODO: We should be able to configure IO chunks based on size.
// distribute utterances over chunks
// We simply count off frames until we reach the chunk size.
// Note that we first randomize the chunks, i.e. when used, chunks are non-consecutive and thus cause the disk head to seek for each chunk.
// We have 100 frames in a second.
const size_t FramesPerSec = 100;
// A chunk constitutes of 15 minutes
const size_t ChunkFrames = 15 * 60 * FramesPerSec; // number of frames to target for each chunk
m_chunks.resize(0);
m_chunks.reserve(m_totalNumberOfFrames / ChunkFrames);
ChunkIdType chunkId = 0;
foreach_index(i, utterances)
{
// if exceeding current entry--create a new one
// I.e. our chunks are a little larger than wanted (on av. half the av. utterance length).
if (m_chunks.empty() || m_chunks.back().GetTotalFrames() > ChunkFrames)
{
m_chunks.push_back(HTKChunkDescription(chunkId++));
}
// append utterance to last chunk
HTKChunkDescription& currentChunk = m_chunks.back();
if (!m_primary)
{
// Have to store key <-> utterance mapping for non primary deserializers.
m_keyToChunkLocation[utterances[i].GetId()] = make_pair(currentChunk.GetChunkId(), currentChunk.GetNumberOfUtterances());
}
currentChunk.Add(move(utterances[i]));
}
fprintf(stderr,
"HTKDataDeserializer::HTKDataDeserializer: "
"selected %" PRIu64 " utterances grouped into %" PRIu64 " chunks, "
"average chunk size: %.1f utterances, %.1f frames "
"(for I/O: %.1f utterances, %.1f frames)\n",
utterances.size(),
m_chunks.size(),
utterances.size() / (double)m_chunks.size(),
m_totalNumberOfFrames / (double)m_chunks.size(),
allUtterances / (double)m_chunks.size(),
allFrames / (double)m_chunks.size());
if (utterances.empty())
{
RuntimeError("HTKDataDeserializer: No utterances to process.");
}
}
// Describes exposed stream - a single stream of htk features.
void HTKDataDeserializer::InitializeStreams(const wstring& featureName)
{
StreamDescriptionPtr stream = make_shared<StreamDescription>();
stream->m_id = 0;
stream->m_name = featureName;
stream->m_sampleLayout = make_shared<TensorShape>(m_dimension);
stream->m_elementType = m_elementType;
stream->m_storageType = StorageType::dense;
m_streams.push_back(stream);
}
// Reading information about the features from the first file.
// This information is used later to check that all features among all files have the same properties.
void HTKDataDeserializer::InitializeFeatureInformation()
{
msra::util::attempt(5, [&]()
{
msra::asr::htkfeatreader reader;
reader.getinfo(m_chunks.front().GetUtterance(0)->GetPath(), m_featureKind, m_ioFeatureDimension, m_samplePeriod);
fprintf(stderr, "HTKDataDeserializer::HTKDataDeserializer: determined feature kind as %d-dimensional '%s' with frame shift %.1f ms\n",
(int)m_ioFeatureDimension, m_featureKind.c_str(), m_samplePeriod / 1e4);
});
}
// Gets information about available chunks.
ChunkDescriptions HTKDataDeserializer::GetChunkDescriptions()
{
ChunkDescriptions chunks;
chunks.reserve(m_chunks.size());
for (ChunkIdType i = 0; i < m_chunks.size(); ++i)
{
auto cd = make_shared<ChunkDescription>();
cd->m_id = i;
cd->m_numberOfSamples = m_chunks[i].GetTotalFrames();
// In frame mode, each frame is represented as sequence.
// The augmentation is still done for frames in the same sequence only, please see GetSequenceById method.
cd->m_numberOfSequences = m_frameMode ? m_chunks[i].GetTotalFrames() : m_chunks[i].GetNumberOfUtterances();
chunks.push_back(cd);
}
return chunks;
}
// Gets sequences for a particular chunk.
// This information is used by the randomizer to fill in current windows of sequences.
void HTKDataDeserializer::GetSequencesForChunk(ChunkIdType chunkId, vector<SequenceDescription>& result)
{
const HTKChunkDescription& chunk = m_chunks[chunkId];
result.reserve(m_frameMode ? chunk.GetTotalFrames() : chunk.GetNumberOfUtterances());
size_t offsetInChunk = 0;
for (size_t i = 0; i < chunk.GetNumberOfUtterances(); ++i)
{
auto utterance = chunk.GetUtterance(i);
// Currently we do not support common prefix, so simply assign the minor to the key.
size_t sequence = utterance->GetId();
if (m_frameMode)
{
// Because it is a frame mode, creating a sequence for each frame.
for (size_t k = 0; k < utterance->GetNumberOfFrames(); ++k)
{
SequenceDescription f;
f.m_chunkId = chunkId;
f.m_key.m_sequence = sequence;
f.m_key.m_sample = k;
f.m_id = offsetInChunk++;
f.m_numberOfSamples = 1;
result.push_back(f);
}
}
else
{
// Creating sequence description per utterance.
SequenceDescription f;
f.m_chunkId = chunkId;
f.m_key.m_sequence = sequence;
f.m_key.m_sample = 0;
f.m_id = offsetInChunk++;
if (SEQUENCELEN_MAX < utterance->GetNumberOfFrames())
{
RuntimeError("Maximum number of samples per sequence exceeded");
}
f.m_numberOfSamples = (uint32_t) utterance->GetNumberOfFrames();
result.push_back(f);
}
}
}
// A wrapper around a matrix that views it as a vector of column vectors.
// Does not have any memory associated.
class MatrixAsVectorOfVectors
{
public:
MatrixAsVectorOfVectors(msra::dbn::matrixbase& m)
: m_matrix(m)
{
}
size_t size() const
{
return m_matrix.cols();
}
const_array_ref<float> operator[](size_t j) const
{
return array_ref<float>(&m_matrix(0, j), m_matrix.rows());
}
private:
DISABLE_COPY_AND_MOVE(MatrixAsVectorOfVectors);
msra::dbn::matrixbase& m_matrix;
};
// Represents a chunk data in memory. Given up to the randomizer.
// It is up to the randomizer to decide when to release a particular chunk.
class HTKDataDeserializer::HTKChunk : public Chunk
{
public:
HTKChunk(HTKDataDeserializer* parent, ChunkIdType chunkId) : m_parent(parent), m_chunkId(chunkId)
{
auto& chunkDescription = m_parent->m_chunks[chunkId];
// possibly distributed read
// making several attempts
msra::util::attempt(5, [&]()
{
chunkDescription.RequireData(m_parent->m_featureKind, m_parent->m_ioFeatureDimension, m_parent->m_samplePeriod, m_parent->m_verbosity);
});
}
// Gets data for the sequence.
virtual void GetSequence(size_t sequenceId, vector<SequenceDataPtr>& result) override
{
m_parent->GetSequenceById(m_chunkId, sequenceId, result);
}
// Unloads the data from memory.
~HTKChunk()
{
auto& chunkDescription = m_parent->m_chunks[m_chunkId];
chunkDescription.ReleaseData(m_parent->m_verbosity);
}
private:
DISABLE_COPY_AND_MOVE(HTKChunk);
HTKDataDeserializer* m_parent;
ChunkIdType m_chunkId;
};
// Gets a data chunk with the specified chunk id.
ChunkPtr HTKDataDeserializer::GetChunk(ChunkIdType chunkId)
{
return make_shared<HTKChunk>(this, chunkId);
};
// A matrix that stores all samples of a sequence without padding (differently from ssematrix).
// The number of columns equals the number of samples in the sequence.
// The number of rows equals the size of the feature vector of a sample (= dimensions).
class FeatureMatrix
{
public:
FeatureMatrix(size_t numRows, size_t numColumns) : m_numRows(numRows), m_numColumns(numColumns)
{
m_data.resize(m_numRows * m_numColumns);
}
// Returns a reference to the column.
inline array_ref<float> col(size_t column)
{
return array_ref<float>(m_data.data() + m_numRows * column, m_numRows);
}
// Gets pointer to the data.
inline float* GetData()
{
return m_data.data();
}
// Gets the number of columns. It equals the number of samples in the sequence/utterance.
inline size_t GetNumberOfColumns() const
{
return m_numColumns;
}
// Gets total size in elements of stored features.
inline size_t GetTotalSize() const
{
return m_data.size();
}
private:
// Features
std::vector<float> m_data;
// Number of rows = dimension of the feature
size_t m_numRows;
// Number of columns = number of samples in utterance.
size_t m_numColumns;
};
// This class stores sequence data for HTK for floats.
struct HTKFloatSequenceData : DenseSequenceData
{
HTKFloatSequenceData(FeatureMatrix&& data) : m_buffer(data)
{
m_numberOfSamples = (uint32_t)data.GetNumberOfColumns();
if (m_numberOfSamples != data.GetNumberOfColumns())
{
RuntimeError("Maximum number of samples per sequence exceeded.");
}
}
const void* GetDataBuffer() override
{
return m_buffer.GetData();
}
private:
FeatureMatrix m_buffer;
};
// This class stores sequence data for HTK for doubles.
struct HTKDoubleSequenceData : DenseSequenceData
{
HTKDoubleSequenceData(FeatureMatrix& data) : m_buffer(data.GetData(), data.GetData() + data.GetTotalSize())
{
m_numberOfSamples = (uint32_t)data.GetNumberOfColumns();
if (m_numberOfSamples != data.GetNumberOfColumns())
{
RuntimeError("Maximum number of samples per sequence exceeded.");
}
}
const void* GetDataBuffer() override
{
return m_buffer.data();
}
private:
std::vector<double> m_buffer;
};
// Copies a source into a destination with the specified destination offset.
static void CopyToOffset(const const_array_ref<float>& source, array_ref<float>& destination, size_t offset)
{
size_t sourceSize = source.size() * sizeof(float);
memcpy_s((char*)destination.begin() + sourceSize * offset, sourceSize, &source.front(), sourceSize);
}
// TODO: Move augmentation to the separate class outside of deserializer.
// TODO: Check the CNTK Book why different left and right extents are not supported.
// Augments a frame with a given index with frames to the left and right of it.
static void AugmentNeighbors(const MatrixAsVectorOfVectors& utterance,
size_t frameIndex,
const size_t leftExtent,
const size_t rightExtent,
array_ref<float>& destination)
{
CopyToOffset(utterance[frameIndex], destination, leftExtent);
for (size_t currentFrame = frameIndex, n = 1; n <= leftExtent; n++)
{
if (currentFrame > 0)
currentFrame--; // index does not move beyond boundary
CopyToOffset(utterance[currentFrame], destination, leftExtent - n);
}
for (size_t currentFrame = frameIndex, n = 1; n <= rightExtent; n++)
{
if (currentFrame + 1 < utterance.size())
currentFrame++; // index does not move beyond boundary
CopyToOffset(utterance[currentFrame], destination, leftExtent + n);
}
}
// Get a sequence by its chunk id and sequence id.
// Sequence ids are guaranteed to be unique inside a chunk.
void HTKDataDeserializer::GetSequenceById(ChunkIdType chunkId, size_t id, vector<SequenceDataPtr>& r)
{
const auto& chunkDescription = m_chunks[chunkId];
size_t utteranceIndex = m_frameMode ? chunkDescription.GetUtteranceForChunkFrameIndex(id) : id;
const UtteranceDescription* utterance = chunkDescription.GetUtterance(utteranceIndex);
auto utteranceFrames = chunkDescription.GetUtteranceFrames(utteranceIndex);
// wrapper that allows m[j].size() and m[j][i] as required by augmentneighbors()
MatrixAsVectorOfVectors utteranceFramesWrapper(utteranceFrames);
size_t utteranceLength = m_frameMode ? 1 : (m_expandToPrimary ? utterance->GetExpansionLength() : utterance->GetNumberOfFrames());
FeatureMatrix features(m_dimension, utteranceLength);
if (m_frameMode)
{
// For frame mode augment a single frame.
size_t frameIndex = id - chunkDescription.GetStartFrameIndexInsideChunk(utteranceIndex);
auto fillIn = features.col(0);
AugmentNeighbors(utteranceFramesWrapper, frameIndex, m_augmentationWindow.first, m_augmentationWindow.second, fillIn);
}
else if (m_expandToPrimary) // Broadcast a single frame to the complete utterance.
{
for (size_t resultingIndex = 0; resultingIndex < utterance->GetExpansionLength(); ++resultingIndex)
{
auto fillIn = features.col(resultingIndex);
AugmentNeighbors(utteranceFramesWrapper, 0, m_augmentationWindow.first, m_augmentationWindow.second, fillIn);
}
}
else // Augment the complete utterance.
{
for (size_t frameIndex = 0; frameIndex < utterance->GetNumberOfFrames(); ++frameIndex)
{
auto fillIn = features.col(frameIndex);
AugmentNeighbors(utteranceFramesWrapper, frameIndex, m_augmentationWindow.first, m_augmentationWindow.second, fillIn);
}
}
// Copy features to the sequence depending on the type.
DenseSequenceDataPtr result;
if (m_elementType == ElementType::tdouble)
{
result = make_shared<HTKDoubleSequenceData>(features);
}
else if (m_elementType == ElementType::tfloat)
{
result = make_shared<HTKFloatSequenceData>(std::move(features));
}
else
{
LogicError("Currently, HTK Deserializer supports only double and float types.");
}
r.push_back(result);
}
// Gets sequence description by its key.
bool HTKDataDeserializer::GetSequenceDescription(const SequenceDescription& primary, SequenceDescription& d)
{
assert(!m_primary);
auto iter = m_keyToChunkLocation.find(primary.m_key.m_sequence);
if (iter == m_keyToChunkLocation.end())
{
return false;
}
auto chunkId = iter->second.first;
auto utteranceIndexInsideChunk = iter->second.second;
auto& chunk = m_chunks[chunkId];
auto utterance = chunk.GetUtterance(utteranceIndexInsideChunk);
d.m_chunkId = (ChunkIdType)chunkId;
// TODO: When we move frame mode from deserializer, expanding should go away and be done on the higher level.
// TODO: Currently for the frame mode the secondary deserializer does not know the size of the full utterance,
// becase each frame has its own sequence description. So we get the length by the max sample we have seen.
if (m_expandToPrimary)
{
// Expanding for sequence length/or max seen frame.
size_t maxLength = max(primary.m_numberOfSamples, (uint32_t)primary.m_key.m_sample + 1);
if (utterance->GetExpansionLength() < maxLength)
{
utterance->SetExpansionLength(maxLength);
}
d.m_id = utteranceIndexInsideChunk;
}
else
{
d.m_id = m_frameMode ? chunk.GetStartFrameIndexInsideChunk(utteranceIndexInsideChunk) + primary.m_key.m_sample : utteranceIndexInsideChunk;
}
d.m_numberOfSamples = m_frameMode ? 1 : (uint32_t)utterance->GetNumberOfFrames();
return true;
}
}}}