Views
No views yet
SynthVoice: This should be a Research Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦 Repo We release the raw data for our processed LibriVoice dataset, derived from the original LibriSpeech dataset maintained by OpenSLR. The data format for each sample in the voice_segments.jsonl is as follows: json { "speaker_id": "", "audio_path": "", "transcript": "", "duration_ms": "", "sample_rate": 16000, "source": "", "quality_score": "" } Some samples may have incomplete… See the full description on the dataset page: https://huggingface.co/datasets/toolevalxm/Synth-LibriVoice-Raw.