Datasetpublic

SynDataLab-EN/tts-pretrain-clones-3m-mos

TTS Pretrain Clones (3M) — with DNSMOS This is SynDataLab/tts-pretrain-clones-3m with an added per-utterance dnsmos column (DNSMOS P.835 OVRL score, float32), computed with the sig_bak_ovr.onnx model. 2,967,779 clone utterances across 2971 English speakers. Sample rate: 44.1 kHz, WAV in Parquet dnsmos: overall MOS quality estimate per utterance (higher is better) Generated by echo-tts synthesizing English text on speaker latents derived from Qwen3-TTS VoiceDesign base speakers.… See the full description on the dataset page: https://huggingface.co/datasets/SynDataLab-EN/tts-pretrain-clones-3m-mos.

sourceHugging Faceapache-2.0updated 4mo agoView on Hugging Face

0likes830downloads

Dataset Card

TTS Pretrain Clones (3M) — with DNSMOS

This is SynDataLab/tts-pretrain-clones-3m with an added per-utterance `dnsmos` column (DNSMOS P.835 OVRL score, float32), computed with the sig_bak_ovr.onnx model.

2,967,779 clone utterances across 2971 English speakers.

—Sample rate: 44.1 kHz, WAV in Parquet
—dnsmos: overall MOS quality estimate per utterance (higher is better)
—Generated by echo-tts synthesizing English text on speaker latents derived from Qwen3-TTS VoiceDesign base speakers. Per speaker: 10 voice-clone latents × 100 texts.
—Coverage: speakers 1-60 + 61 (partial, 749 rows) + 91-3000.