|
|
--- |
|
|
license: cc-by-sa-4.0 |
|
|
configs: |
|
|
- config_name: default |
|
|
data_files: |
|
|
- split: train |
|
|
path: data/train-* |
|
|
- split: validation |
|
|
path: data/validation-* |
|
|
dataset_info: |
|
|
features: |
|
|
- name: Message |
|
|
dtype: string |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 171194856 |
|
|
num_examples: 8923544 |
|
|
- name: validation |
|
|
num_bytes: 943300 |
|
|
num_examples: 61113 |
|
|
download_size: 98397571 |
|
|
dataset_size: 172138156 |
|
|
task_categories: |
|
|
- text-classification |
|
|
- text-generation |
|
|
language: |
|
|
- en |
|
|
tags: |
|
|
- twitch |
|
|
- y2k |
|
|
- chat |
|
|
- genz |
|
|
pretty_name: twitch |
|
|
size_categories: |
|
|
- 1M<n<10M |
|
|
--- |
|
|
|
|
|
# Twitch Chat Dataset |
|
|
 |
|
|
|
|
|
|
|
|
This dataset is a large-scale collection of Twitch chat logs aggregated from multiple streamers across various categories. It is designed to support the research and development of models for real-time, informal, and community-driven conversation, such as: |
|
|
|
|
|
- Chatbots tailored for livestream platforms |
|
|
- Simulating the behavior of Twitch chat |
|
|
- Modeling how chat reacts during hype moments, events, or memes |
|
|
|
|
|
The code for it can be found [here](https://gist.github.com/Parkourer10/1a563c7359e786b40f8bf600d8d04573) |
|
|
|
|
|
--- |
|
|
|
|
|
## 📂 Dataset Structure |
|
|
|
|
|
- All files are combined into the **`train`** split, except the last one, which is used as the **`validation`** split. |
|
|
- The `message` column contains raw Twitch chat messages. |
|
|
|
|
|
--- |
|
|
|
|
|
## 💡 Use Cases |
|
|
|
|
|
- Fine-tuning language models (e.g., LLaMA, Mistral, GPT) on informal, emoji-heavy, meme-rich dialogue. |
|
|
- Building generative agents or NPCs that mimic Twitch chat behavior. |
|
|
- Training moderation, toxicity, or sentiment classifiers for real-world fast-paced chat. |
|
|
- Analyzing the dynamics of online communities, virality, and meme propagation. |
|
|
|
|
|
--- |
|
|
|
|
|
## 🔒 Ethics |
|
|
|
|
|
- **Usernames are removed** to protect individual privacy. |
|
|
- **No private messages, subscriptions, or off-platform data** are included. |
|
|
If you find any such data, please open a discussion so it can be removed. |
|
|
- Always review and comply with [Twitch's Terms of Service](https://www.twitch.tv/p/en/legal/terms-of-service/) when using, modifying, or deploying models trained on this dataset. |
|
|
|
|
|
--- |
|
|
|
|
|
## 📜 License |
|
|
|
|
|
This dataset is licensed under **[CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0/)**. |
|
|
You may use it for research or commercial purposes, provided that you give appropriate credit and share any derivative works under the same license. |
|
|
|
|
|
--- |
|
|
|
|
|
## 🙏 Acknowledgements |
|
|
|
|
|
Special thanks to the Twitch community and to the streamers whose public broadcasts and chat logs made this dataset possible. |
|
|
|
|
|
--- |
|
|
|
|
|
## 🔗 Citation |
|
|
|
|
|
If you use this dataset in your research or applications, please cite it as: |
|
|
|
|
|
```bibtex |
|
|
@misc{twitchchat2025, |
|
|
title = {Twitch Chat}, |
|
|
author = {parkourer10}, |
|
|
year = {2025}, |
|
|
publisher = {Hugging Face Datasets}, |
|
|
url = {https://huggingface.co/datasets/lparkourer10/twitch_chat} |
|
|
} |
