punchline
punchline
Bangumi Image Base of Punch Line
This is the image base of bangumi Punch Line, we detected 17 characters, 1203 images in total. The full dataset is here.
Please note that these image bases are not guaranteed to be 100% cleaned, they may be noisy actual. If you intend to manually train models using this dataset, we recommend performing necessary preprocessing on the downloaded dataset to eliminate potential noisy samples (approximately 1% probability).
Here is the characters'… See the full description on the dataset page: https://huggingface.co/datasets/BangumiBase/punchline.ruozhiba-punchline我上个项目的进阶版，用来提升LLM的幽默感。
数据来自github
调用GPT将弱智吧query改写成问答对形式，并将punchiline笑点设置在answer部分。
额外添加CoT模块，但CoT部分错误率较高，经验上不建议加入训练。
基于该数据集训练模型的推理结果show case如下
Q: 欲望和收入不匹配这件事情你怎么看
A: 我觉得只要心态好，慢慢挣就好了，欲速则布达拉宫
Q：握不住的沙就扬了吧
A：阿sir，这种话让人听了很不舒服，建议不要说出来。
qwantz-punchline-classificationshort-jokes-punchline
Short Jokes Punchline
This dataset contains information about jokes, visitors, labels, and label segments used in a joke labeling application. The data is stored in four CSV files: joke.csv, visitor.csv, label.csv, and label_segment.csv.
Files
joke.csv
This file contains 200 jokes randomly sampled from the Kaggle dataset "Short Jokes." Each row represents a joke with the following columns:
id: The unique identifier for the joke.
text: The text content of the… See the full description on the dataset page: https://huggingface.co/datasets/Timxjl/short-jokes-punchline.ruozhiba-punchline我上个项目的进阶版，用来提升LLM的幽默感。
数据来自github
调用GPT将弱智吧query改写成问答对形式，并将punchiline笑点设置在answer部分。
额外添加CoT模块，但CoT部分错误率较高，经验上不建议加入训练。
基于该数据集训练模型的推理结果show case如下
Q: 欲望和收入不匹配这件事情你怎么看
A: 我觉得只要心态好，慢慢挣就好了，欲速则布达拉宫
Q：握不住的沙就扬了吧
A：阿sir，这种话让人听了很不舒服，建议不要说出来。
ruozhiba_punchline_ft