tabisheva / streaming-transformer

Streaming transformer

Pytorch implementation of Augmented Memory Transformer for streaming automatic speech recognition with linear attention mechanism from this paper.

Data preparation

Download LJSpeech dataset

cd data
wget https://data.keithito.com/data/speech/LJSpeech-1.1.tar.bz2   # download data 
tar xjf LJSpeech-1.1.tar.bz2                                      # extract data
python prepare_vocabulary.py                                      # building target dictionary

Training

python train.py