Tag
Author highlights under-discussed text normalization issues in streaming TTS and shares a vendor benchmark evaluating 1000+ sentences across 31 categories for dates, URLs, acronyms, etc.
easyaligner is an open-source forced alignment library with GPU acceleration and flexible text normalization that works with all wav2vec2 models on Hugging Face Hub. It addresses practical workflows like handling partial transcripts, irrelevant speech segments, and long audio without chunking while preserving original text formatting.