ReasoningV-5K
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/BUAA-CLab/ReasoningV
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个高质量的数据集，包含5,000个经过功能验证的实例，这些实例通过多维筛选PyraNet样本创建了推理路径。该数据集旨在提升大型语言模型在Verilog代码生成任务中的推理能力。数据集的规模为5,000个实例，其任务是专注于Verilog代码生成。
This is a high-quality dataset containing 5,000 functionally validated instances, for which reasoning paths were created via multi-dimensional screening of PyraNet samples. This dataset aims to enhance the reasoning capabilities of large language models (LLMs) in Verilog code generation tasks. The dataset consists of 5,000 instances, with its core task focused on Verilog code generation.
提供机构：
BUAA-CLab
