metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: transcription
      dtype: string
    - name: duration
      dtype: float64
  splits:
    - name: train
      num_bytes: 52713823
      num_examples: 1566
    - name: test
      num_bytes: 2744624
      num_examples: 83
  download_size: 55463391
  dataset_size: 55458447
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
license: apache-2.0
language:
  - crh
task_categories:
  - text-to-speech
size_categories:
  - 1K<n<10K

Open Source Crimean Tatar Text-to-Speech datasets

This is subset of Sevil voice with train/test splits.

Community

Discord: https://bit.ly/discord-uds
Speech Recognition: https://t.me/speech_recognition_uk
Speech Synthesis: https://t.me/speech_synthesis_uk

Statistic

Quality: high
Duration: 2h29m
Frequency: 48 kHz

Cite this work

@misc {smoliakov_2025,
    author       = { {Smoliakov} },
    title        = { qirimtatar-tts (Revision c2ceec6) },
    year         = 2025,
    url          = { https://huggingface.co/datasets/Yehor/qirimtatar-tts },
    doi          = { 10.57967/hf/4608 },
    publisher    = { Hugging Face }
}