Classes | |
class | Alignment |
represents the core of attention mechanism More... | |
class | Attention |
calculates attention vector from context and decoder output vectors More... | |
class | BeamSearchPolicy |
processes the results of one iteration of the generator with beam search and produces input for the next iteration More... | |
class | BenchmarkWriter |
all it does is to measure the performance of sequence generation More... | |
class | BLEUScoreWriter |
all it does is to evaluate BLEU score More... | |
class | Component |
a functional part of the sample More... | |
class | ComponentWeights |
weights storage More... | |
class | Context |
calculates context vector from raw alignment scores and memory states More... | |
class | DataReader |
reader of sequences of data More... | |
class | DataWriter |
writer of sequences of data More... | |
class | DebugUtil |
container for static debug utility functions More... | |
class | Decoder |
encodes single input into output states More... | |
class | DeviceBuffer |
class | Embedder |
projects 1-hot vectors (represented as a vector with indices) into dense embedding space More... | |
class | Encoder |
encodes input sentences into output states More... | |
class | Likelihood |
calculates likelihood and TopK indices for the raw input logits More... | |
class | LikelihoodCombinationOperator |
class | LimitedSamplesDataReader |
wraps another data reader and limits the number of samples to read More... | |
class | LSTMDecoder |
encodes single input into output states with LSTM More... | |
class | LSTMEncoder |
encodes input sentences into output states using LSTM More... | |
class | MultiplicativeAlignment |
alignment scores from Luong attention mechanism More... | |
class | PinnedHostBuffer |
wrapper for the pinned host memory region More... | |
class | Projection |
calculates raw logits More... | |
class | SequenceProperties |
provides encoder/decoder relevant properties of sequences More... | |
class | SLPAttention |
Linear attention calculation. More... | |
class | SLPEmbedder |
selects the embedding vector from the weight matrix using index provided in the input More... | |
class | SLPProjection |
Linear logits calculation. More... | |
class | SoftmaxLikelihood |
calculates softmax likelihood and TopK indices for the raw input logits More... | |
class | TextReader |
reads sequences of data from input stream More... | |
class | TextWriter |
class | Vocabulary |
String<->Id bijection storage. More... | |
Typedefs | |
typedef std::vector< std::string > | Segment_t |
typedef std::map< Segment_t, int > | Count_t |
Functions | |
int | read (std::vector< Segment_t > &samples, std::shared_ptr< std::istream > input, int samplesToRead=1) |
Count_t | ngramCounts (const Segment_t &segment, int maxOrder=4) |
Count_t | ngramCountIntersection (const Count_t &cnt0, const Count_t &cnt1) |
void | accumulateBLEU (const std::vector< Segment_t > &referenceSamples, const std::vector< Segment_t > &outputSamples, int maxOrder, size_t &referenceLength, size_t &translationLength, std::vector< size_t > &matchesByOrder, std::vector< size_t > &possibleMatchesByOrder) |
std::istream & | operator>> (std::istream &input, Vocabulary &value) |
std::istream & | operator>> (std::istream &input, ComponentWeights &value) |
int | inferTypeToBytes (nvinfer1::DataType t) |
int | getVolume (nvinfer1::Dims dims) |
std::vector< float > | resizeWeights (int rows, int cols, int rowsNew, int colsNew, const float *memory) |
typedef std::vector<std::string> nmtSample::Segment_t |
typedef std::map<Segment_t, int> nmtSample::Count_t |
int nmtSample::read | ( | std::vector< Segment_t > & | samples, |
std::shared_ptr< std::istream > | input, | ||
int | samplesToRead = 1 |
||
) |
void nmtSample::accumulateBLEU | ( | const std::vector< Segment_t > & | referenceSamples, |
const std::vector< Segment_t > & | outputSamples, | ||
int | maxOrder, | ||
size_t & | referenceLength, | ||
size_t & | translationLength, | ||
std::vector< size_t > & | matchesByOrder, | ||
std::vector< size_t > & | possibleMatchesByOrder | ||
) |
std::istream& nmtSample::operator>> | ( | std::istream & | input, |
Vocabulary & | value | ||
) |
std::istream& nmtSample::operator>> | ( | std::istream & | input, |
ComponentWeights & | value | ||
) |
int nmtSample::inferTypeToBytes | ( | nvinfer1::DataType | t | ) |
int nmtSample::getVolume | ( | nvinfer1::Dims | dims | ) |