Text-to-Speech on LibriTTS (Unseen)
2.31WERMaskGCT
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MaskGCTParams=315M, Data (h)=100K2026.06 | 2.31 | 3.74 | 0.156 | 1.071 | 0.9495 | |
| CosyVoiceParams=300M, Data (h)=170K2026.06 | 2.46 | 3.95 | 0.152 | 1.06 | 0.9117 | |
| S5-TTS w/ IMSDParams=160M, Data (h)=4.67K2026.06 | 2.65 | 3.72 | 0.179 | 1.075 | 0.934 | |
| E2-TTSParams=335M, Data (h)=100K2026.06 | 2.82 | 3.65 | 0.137 | 1.071 | 0.9487 | |
| FireRedTTSParams=400M, Data (h)=248K2026.06 | 4.7 | 3.82 | 0.157 | 1.074 | 0.9229 |