datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
arabic-tts-wav-24k
Arabic TTS WAV 24k Dataset
A high-quality, open-source dataset for Arabic Text-to-Speech (TTS) research, containing paired audio and text samples from both male and female speakers. All audio is provided in 24kHz WAV format, with rich metadata and phonetic transcriptions.
Dataset Summary
This dataset is designed for training and evaluating neural TTS systems in Modern Standard Arabic. It includes:
Audio: Clean, studio-quality WAV files at 24,000 Hz.
Text: Original Arabic… See the full description on the dataset page: https://huggingface.co/datasets/NeoBoy/arabic-tts-wav-24k.DART-DATASETSNezoxelevenlabsSpeechTest
ElevenLabs Speech Dataset
This dataset contains speech data generated using the ElevenLabs API. It includes phrases in various variants, processed with different stability settings and recorded by multiple speakers.
Data Description:
Transcriptions: Each transcription corresponds to a phrase that was spoken by one of 10 different speakers.
Speakers: 10 different speakers were used.
Stability Levels: 5 stability levels were applied to each transcription.
Variants: The… See the full description on the dataset page: https://huggingface.co/datasets/NeoBoy/elevenlabsSpeechTest.sdr-sigwiki-jointtrelis_voice_512_V5sdr-sigwiki-audioNEONtrelis_voice_512_V2vozdericovozrodolfotestevoztrelis_voice_512trelis_voice_512_V3