glaswegian_audio / README.md
divakaivan's picture
Update README.md
ee2714c verified
---
license: mit
task_categories:
- text-to-speech
language:
- en
tags:
- glaswegian
- scottish
size_categories:
- n<1K
---
### WARNING! Some derogatory slang is included in the dataset
Latest total length: 120 minutes
Source:
[Scottish phrases 1-10](https://www.youtube.com/@Cannaekickaroonhere/videos), privately recorded audio, [Limmy](https://www.youtube.com/@DaftLimmy), and 1 episode of [Glasga Da](https://youtu.be/CRM_npynkvA)
Metadata:
| Column Name | Data Type | Information |
|----------------------|----------|-----------------------------------------------------------------------------------------------|
| index | int | Unique identifier for each row |
| file_name | string | Path to the audio file |
| transcription | string | Text transcription of the audio |
| length_seconds | float | Length of the audio file in seconds |
| sampling_rate | int | Sampling rate of the audio file, typically in Hertz (Hz) |
| gender | string | Gender of the speaker |
| age | string | Age range of the speaker |
| class | string | Socioeconomic class or demographic classification of the speaker |
| location | string | Geographic origin or region of the speaker |
| speaker_id | int | Unique identifier for the speaker, useful for distinguishing between different speakers |
Fine-tuned whisper-small using this dataset:
* [model](https://huggingface.co/divakaivan/glaswegian-asr)
* [demo](https://huggingface.co/spaces/divakaivan/glaswegian-whisper)
Fine-tuned TTS using this dataset:
* [model](https://huggingface.co/divakaivan/glaswegian_tts)
* [demo](https://huggingface.co/spaces/divakaivan/glaswegian-tts-demo)
### Contact
If you have any questions related to the dataset or models, please contact the authors on LinkedIn:
* [Colin McKelvie](https://www.linkedin.com/in/colin-mckelvie-998605279/)
* [Ivan Ivanov](https://www.linkedin.com/in/ivansivanov/)