JingweiNi/train_prm800k_qwen3_8b_thinking_5000_extracted_1
收藏Hugging Face2025-12-15 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/JingweiNi/train_prm800k_qwen3_8b_thinking_5000_extracted_1
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个特征，包括question（问题）、answer（答案）、input_ids（输入ID）、reply（回复）、original_index（原始索引）和claims（声明），其中claims是一个列表，包含aligned_token_ids（对齐的令牌ID）、claim_text（声明文本）和sentence（句子）。数据集只有一个名为train的分割，包含625个示例，总大小为53989970字节。该数据集可能用于涉及问题、答案和声明的自然语言处理任务，如问答或声明验证。
The dataset includes features such as question, answer, input_ids, reply, original_index, and claims, where claims is a list containing aligned_token_ids, claim_text, and sentence. The dataset has a single split named train with 625 examples and a total size of 53989970 bytes. It is likely designed for NLP tasks involving questions, answers, and claims, such as question answering or claim verification.
提供机构：
JingweiNi
