Types & classes — github.com/HawkAaron/RNN-Transducer

CodeHub MCPcopy Create free account

hub / github.com/HawkAaron/RNN-Transducer / types & classes

Types & classes17 in github.com/HawkAaron/RNN-Transducer

⨍ Functions 82 ◇ Types & classes 17

↓ 8 callersClassSequentialLoader

DataLoader.py:38

↓ 2 callersClassRNNModel

A model with an encoder, recurrent layer, and a decoder.

↓ 2 callersClassRNNModel

A model with an encoder, recurrent layer, and a decoder.

↓ 2 callersClassRNNTLoss

↓ 2 callersClassSequence

model2012.py:162

↓ 2 callersClassSequence

↓ 1 callersClassAttention

seq2seq/attention.py:4

↓ 1 callersClassCTCLoss

train_ctc.py:13

↓ 1 callersClassDecoder

seq2seq/decoder.py:7

↓ 1 callersClassEncoder

seq2seq/encoder.py:5

↓ 1 callersClassRNNTLoss

↓ 1 callersClassRNNTransducer

The implementation of RNN Transducer loss functions. To make it usable for real-world cases, this class has two policies below. 1. This class

↓ 1 callersClassSeq2seq

seq2seq/seq2seq.py:10

↓ 1 callersClassTransducer

When joint training, remove RNNModel decoder layer

ClassRNNTransducerProp

DataLoader.py:78

ClassTransducer

When joint training, remove RNNModel decoder layer

model2012.py:25