--- language: - uk license: apache-2.0 task_categories: - text-to-speech --- # πŸ‡ΊπŸ‡¦ Open Source Ukrainian Text-to-Speech dataset named LADA Join Ukrainian community - https://t.me/speech_synthesis_uk More details about this dataset - https://github.com/egorsmkv/ukrainian-tts-datasets/tree/main/lada # Voice LADA (female) License: [Apache 2.0](https://github.com/egorsmkv/ukrainian-tts-datasets/blob/main/LICENSE) **Samples are manually checked. Good samples are in the `accept` folder (10h37m), others are in `reject` (1h2m).** Listen to [DEMO](https://huggingface.co/spaces/theodotus/ukrainian-voices) (choose "lada" in the Voice field) ## Features - Quality: high - Duration: 10h37m - Audio formats: OPUS/WAV - Text format: JSONL (a `metadata.jsonl` file) - Frequency: 16000/22050/48000 Hz ## Original version ### In the `OPUS` format - 48000 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_ogg.zip ### In the `WAV` format - 48000 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_48khz.zip - 22050 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_22khz.zip - 16000 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_16khz.zip ## Trimmed version (removed silence) Silence is removed by https://github.com/proger/uk#align-text-to-audio-and-trim-silence ### In the `WAV` format - 48000 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_trimmed_48khz.zip - 22050 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_trimmed_22khz.zip - 16000 Hz: https://huggingface.co/datasets/Yehor/ukrainian-tts-lada/resolve/main/dataset_lada_trimmed_16khz.zip