- Addresses phoneme generation bug mentioned in 5 - Longform inference supported (previously limited to ~450 character inputs) - Visit [PyPi project](https://pypi.org/project/styletts2/) for earlier releases
0.1.5
- Option to pass style vector directly to `inference()` - Addresses CVEs from transformers, scipy (this also means Python 3.9 no longer supported from v0.1.5 and beyond) - Visit [PyPi project](https://pypi.org/project/styletts2/) for earlier releases