Contrastive-Disentangled-Memory-for-sequential-Learning

Introduction

This is a source code for Transformer with a contrastive disentangled memory. The implementation includes machine translation and online speech recognition.

Requirements and Frameworks

• Python 3.6.1+
• Fairseq
• ESPnet

To do machine translation

cd Machine_Translation

To do speech recognition

cd Speech_Recognition