Exploring the Effect of Segmentation and Vocabulary Size on Speech Tokenization for Speech Language Models

Open in new window