| --- |
| license: other |
| task_categories: |
| - text-to-speech |
| - automatic-speech-recognition |
| language: |
| - en |
| tags: |
| - audio |
| - tts |
| - voice-acting |
| - slay-the-princess |
| pretty_name: Slay the Princess – Voice Dataset |
| size_categories: |
| - 1K<n<10K |
| dataset_info: |
| features: |
| - name: audio |
| dtype: |
| audio: |
| sampling_rate: 44100 |
| - name: text |
| dtype: string |
| - name: duration |
| dtype: float32 |
| - name: speaker |
| dtype: string |
| splits: |
| - name: train |
| num_bytes: 2239005055 |
| num_examples: 9394 |
| download_size: 2507291079 |
| dataset_size: 2239005055 |
| configs: |
| - config_name: default |
| data_files: |
| - split: train |
| path: data/train-* |
| extra_gated_prompt: >- |
| This dataset contains audio extracted from Slay the Princess, a game by |
| Black Tabby Games. The audio is the intellectual property of Black Tabby Games |
| and Abby Howard. By accessing this dataset you agree to: |
| (1) use it for non-commercial research or personal projects only, |
| (2) not redistribute the audio files, |
| (3) credit Black Tabby Games in any publication or project that uses this data. |
| If you are a rights holder and want this dataset removed, please open a |
| community discussion and it will be taken down promptly. |
| extra_gated_fields: |
| Name: text |
| Affiliation: text |
| Intended use: text |
| I agree to the terms above and will not use this for commercial purposes: checkbox |
| --- |
| |
| # Slay the Princess – Voice Dataset |
|
|
| > **Access is gated.** Fill in the short form above to agree to the non-commercial |
| > research terms before downloading. |
|
|
|
|
| High-quality voice-acted audio lines extracted from |
| [Slay the Princess](https://store.steampowered.com/app/1989270/Slay_the_Princess/) |
| paired with their transcript text. |
|
|
| **Audio format:** FLAC, 44.1 kHz |
| **Total clips:** 9 394 |
| **Total duration:** ~13 h |
|
|
| ## Speakers |
|
|
| | Speaker | Clips | Description | |
| |---|---|---| |
| | narrator | 2 723 | The Narrator | |
| | princess | 2 170 | The Princess (all forms) | |
| | hero | 1 399 | Voice of the Hero | |
| | mound | 469 | The Mound | |
| | stubborn | 345 | Voice of the Stubborn | |
| | skeptic | 332 | Voice of the Skeptic | |
| | opportunist | 331 | Voice of the Opportunist | |
| | cold | 265 | Voice of the Cold | |
| | paranoid | 243 | Voice of the Paranoid | |
| | broken | 246 | Voice of the Broken | |
| | cheated | 240 | Voice of the Cheated | |
| | contrarian | 235 | Voice of the Contrarian | |
| | hunted | 200 | Voice of the Hunted | |
| | smitten | 193 | Voice of the Smitten | |
| | flinching | 2 | Voice of the Flinching | |
|
|
| ## Dataset structure |
|
|
| ```python |
| DatasetDict({ |
| train: Dataset({ |
| features: { |
| 'audio': Audio(sampling_rate=44100), |
| 'text': Value('string'), |
| 'duration': Value('float32'), |
| 'speaker': Value('string'), |
| }, |
| num_rows: 9394 |
| }) |
| }) |
| ``` |
|
|
| ## Usage |
|
|
| ```python |
| from datasets import load_dataset |
| |
| ds = load_dataset("TeraSpace/SlayPrincess", split="train") |
| |
| # Listen to a sample |
| sample = ds[0] |
| print(sample["speaker"], "–", sample["text"]) |
| # sample["audio"]["array"] → numpy array |
| # sample["audio"]["sampling_rate"] → 44100 |
| ``` |
|
|
| ## Copyright |
|
|
| All audio content is © Abby Howard / Black Tabby Games. This dataset is an |
| unofficial fan resource shared strictly for non-commercial research. The |
| maintainer claims no ownership of the source material. To request removal, |
| open a [community discussion](https://huggingface.co/datasets/TeraSpace/SlayPrincess/discussions). |
|
|
