v1.0

• Improved voice quality
• Significantly reduced latency in streaming synthesis
• Reduced model size
• Advanced text normalization

v0.2

• Support for streaming synthesis
• Reduced model size
• Improved performance
• Support for word alignments
• Improved naturalness and pronunciations
• Improved text normalization