Utilizing Character and Word Embeddings for Text Normalization with Sequence-to-Sequence Models

Open in new window