Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models

Devs

Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models | Read Paper on Bytez