fittar/lyric_canvas

---
license: mit
---
# LyricCanvas Dataset
- The lyricCanvas dataset contains approximately 10M lines of lyrics with corresponding visual elaborations (visualizable prompts).
- It could be used to train large language models to translate highly abstract concepts and metaphorical
phrases to visualizable prompts for image generation, see [ViPE](https://huggingface.co/fittar/ViPE-M-CTX7).
- Due to copyright policies, we are not allowed to publish the lyrics, however, we release the visual elaborations and the scraper through which
you can collect the lyrics and rebuild LyricCanvas with no additional cost.

## Compiling LyricCanvas
- Download the lyric_canvas.csv file on this repository
- Follow the steps laid out [here](https://github.com/Hazel1994/ViPE/tree/main/lyric_canvas) to complete the dataset
- Enjoy!

## Citation

If you found LyricCanvas useful, please consider citing

```
@inproceedings{shahmohammadi-etal-2023-vipe,
title = "{V}i{PE}: Visualise Pretty-much Everything",
author = "Shahmohammadi, Hassan and
Ghosh, Adhiraj and
Lensch, Hendrik",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.emnlp-main.333",
pages = "5477--5494"
}
```

原始信息汇总

LyricCanvas 数据集

• 数据集概述：LyricCanvas 数据集包含约 1000 万行歌词及其对应的视觉细化（可视化提示）。
• 用途：该数据集可用于训练大型语言模型，将高度抽象的概念和隐喻性短语转换为可视化提示，用于图像生成。参见 ViPE。
• 版权限制：由于版权政策限制，我们无法发布歌词内容，但我们发布了视觉细化内容和爬虫工具，用户可以通过这些工具免费收集歌词并重建 LyricCanvas 数据集。

编译 LyricCanvas

• 下载文件：从本仓库下载 lyric_canvas.csv 文件。
• 编译步骤：按照 此处 提供的步骤完成数据集的编译。

引用

如果您发现 LyricCanvas 数据集对您的工作有用，请考虑引用以下文献：

@inproceedings{shahmohammadi-etal-2023-vipe, title = "{V}i{PE}: Visualise Pretty-much Everything", author = "Shahmohammadi, Hassan and Ghosh, Adhiraj and Lensch, Hendrik", editor = "Bouamor, Houda and Pino, Juan and Bali, Kalika", booktitle = "Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing", month = dec, year = "2023", address = "Singapore", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2023.emnlp-main.333", pages = "5477--5494" }