CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Models

Benchmark Model Rank Results
text-to-speech-synthesis-on-seed-tts-test-hardCosyVoice 2#6CER: 6.83Speaker Similarity: 0.724
text-to-speech-synthesis-on-seed-tts-test-zhCosyVoice 2-S#15CER: 1.45Speaker Similarity: 0.753
text-to-speech-synthesis-on-seed-tts-test-zhCosyVoice#21CER: 3.63Speaker Similarity: 0.723
text-to-speech-synthesis-on-seed-tts-test-enCosyVoice 2-S#16WER: 2.38Speaker Similarity: 0.654
text-to-speech-synthesis-on-seed-tts-test-enCosyVoice#22WER: 4.29Speaker Similarity: 0.609