| ▼Nnabu | Nabu is a toolbox for designing, training and using end-to-end neural network systems for Automatic Speech Recognition |
| ▼Nneuralnetworks | |
| ▼Ncomponents | |
| ►Nattention | |
| CBahdanauAttention | Normal Bahdanau Style attention |
| CLocationAwareAttention | Attention mechanism that is location aware |
| CWindowedAttention | Attention mechanism that is location aware |
| ►Nbeam_search_decoder | |
| CBeamSearchDecoder | Beam search decode |
| CBeamSearchDecoderFinalOutput | Class for the final output of the BeamSearchDecoder |
| CBeamSearchDecoderOutput | Class for the output of the BeamSearchDecoder |
| CBeamSearchState | Class for the beam search state |
| ►Nconstraints | |
| CMaxNorm | Constraint for a maximum norm of the weights |
| ►Nhooks | |
| CLoadAtBegin | Training hook for saving the final variables |
| CSaveAtEnd | Training hook for saving the final variables |
| CStopHook | Hook that makes sure all replicas terminate when session ends |
| CValidationSaveHook | Training hook for saving and loading the validated model |
| ►Nrnn_cell | |
| CAttentionProjectionWrapper | This wraps an RNN cell to take the context and state and maps it to a fixed diension |
| CBypassWrapper | This will concatenate the input with the output |
| CScopeRNNCellWrapper | This wraps an RNN cell to make sure it uses the same scope every time it's called |
| CStateOutputWrapper | This wraps an RNN cell to make it output its concatenated state instead of the output |
| ▼Ndecoders | |
| ►Nalignment_decoder | |
| CAlignmentDecoder | Gets the HMM state posteriors |
| ►Nbeam_search_decoder | |
| CBeamSearchDecoder | Beam search decode |
| ►Nctc_decoder | |
| CCTCDecoder | CTC Decode |
| ►Ndecoder | |
| CDecoder | Abstract class for a decode |
| ►Nfeature_decoder | |
| CFeatureDecoder | Feature Decode |
| ►Nmax_decoder | |
| CMaxDecoder | Max Decoder takes the maximum probability at each timestep |
| ►Nrandom_decoder | |
| CRandomDecoder | Decoder that returns a random sample from the output distribution |
| ►Nthreshold_decoder | |
| CThresholdDecoder | CTC Decode |
| ▼Nevaluators | |
| ►Ndecoder_evaluator | |
| CDecoderEvaluator | The Decoder Evaluator is used to evaluate a decode |
| ►Nevaluator | |
| CEvaluator | General evaluator class |
| ►Nloss_evaluator | |
| CLossEvaluator | The Decoder Evaluator is used to evaluate a decode |
| ▼Nmodels | |
| ►Ned_decoders | |
| ►Ndnn_decoder | |
| CDNNDecoder | DNN decode |
| ►Ned_decoder | |
| CEDDecoder | General decoder for an encoder decoder system |
| ►Nhotstart_decoder | |
| CHotstartDecoder | Listener object |
| ►Nrnn_decoder | |
| CRNNDecoder | Speller decoder for the LAS architecture |
| ►Nspeller | |
| CSpeller | Speller decoder for the LAS architecture |
| ►Ned_encoders | |
| ►Ndblstm | |
| CDBLSTM | A deep bidirectional LSTM classifie |
| ►Ndnn | |
| CDNN | DNN encode |
| ►Ndummy_encoder | |
| CDummyEncoder | Encoder that does nothing |
| ►Ned_encoder | |
| CEDEncoder | General encoder for an encoder decoder system |
| ►Nhotstart_encoder | |
| CHotstartEncoder | Listener object |
| ►Nlistener | |
| CListener | Listener object |
| ►Nmodel | |
| CModel | General class for an encoder decoder system |
| ▼Nrecognizer | |
| CRecognizer | Recognizer can use a model to produce decode |
| ▼Ntrainers | |
| ►Nstandard_trainer | |
| CStandardTrainer | Trainer with no added functionality |
| ►Ntrainer | |
| CParameterServer | Class for parameter servers |
| CTrainer | General class outlining the training environment of a model |
| ▼Nprocessing | |
| ▼Nprocessors | |
| ►Nalignment_processor | |
| CAlignmentProcessor | Processor for kaldi alignments |
| ►Naudio_processor | |
| CAudioProcessor | Processor for audio files, this will compute features |
| ►Nbinary_processor | |
| CBinaryProcessor | Processor for text data, does normalization |
| ►Nfeature_computers | |
| ►Nfbank | |
| CFbank | Feature computer class to compute fbank features |
| ►Nfeature_computer | |
| CFeatureComputer | A featurecomputer is used to compute features |
| ►Nmfcc | |
| CMfcc | Feature computer class to compute MFCC features |
| ►Nprocessor | |
| CProcessor | General Processor class for data processing |
| ►Ntext_processor | |
| CTextProcessor | Processor for text data, does normalization |
| ►Ntextfile_processor | |
| CTextFileProcessor | Processor for text data where the datalines are pointers to text files, does normalization |
| ▼Ntfreaders | |
| ►Nalignment_reader | |
| CAlignmentReader | Reader for kaldi alignments |
| ►Naudio_feature_reader | |
| CAudioFeatureReader | Reader for audio features |
| ►Nbinary_reader | |
| CBinaryReader | Reader for binary sequences |
| ►Nstring_reader | |
| CStringReader | Reader for reading and encoding text data |
| ►Nstring_reader_eos | |
| CStringReaderEOS | Reader for reading and encoding text data that adds a eos token at the end of every sequence |
| ►Ntfreader | |
| CTfReader | Class for reading tfrecord files and processing them |
| ▼Ntfwriters | |
| ►Nalignment_writer | |
| CAlignmentWriter | TfWriter to write kaldi alignments |
| ►Narray_writer | |
| CArrayWriter | TfWriter to write numpy arrays |
| ►Nbinary_writer | |
| CBinaryWriter | TfWriter to write strings |
| ►Nstring_writer | |
| CStringWriter | TfWriter to write strings |
| ►Ntfwriter | |
| CTfWriter | A class for writing the TF record files |
