barc0/200k_HEAVY_gpt4o-description-gpt4omini-code_generated_problems
收藏Hugging Face2024-11-02 更新2024-12-14 收录
下载链接:
https://hf-mirror.com/datasets/barc0/200k_HEAVY_gpt4o-description-gpt4omini-code_generated_problems
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含约100k的合成数据,这些数据是通过162个种子生成的。生成过程包括使用GPT4o生成描述,然后通过两种方法生成代码,并运行这些代码进行自动过滤,最终得到约200k合法的ARC-like任务及其示例。
This is a dataset of approximately 100,000 synthetic data generated by 162 seeds. The generation process includes generating about 110,000 descriptions using GPT4o, followed by two methods to generate codes: one directly generates codes, and the other suggests using specific library functions to generate codes. These codes are then run and automatically filtered, resulting in approximately 200,000 legitimate ARC-like tasks with examples.
提供机构:
barc0



