TAUR-dev/GPT-4o-mini-eval-DeepSeek-V3_writing-1k
收藏Hugging Face2025-03-10 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/TAUR-dev/GPT-4o-mini-eval-DeepSeek-V3_writing-1k
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个字段，如指令(instruction)、LLM标准(llm_criteria)、数据集名称(dataset)、索引(__index_level_0__)、推理原因(r1_reasoning)、解决方案(r1_solution)、LLM标准拆分(llm_criteria_split)、评估响应(eval_response)、评估解析(eval_parsed)和分数(score)。数据集分为训练集(train)，共有1000个示例，大小为21577501字节。提供了默认配置，并指定了训练数据文件的路径。
The dataset contains multiple fields such as instruction, LLM criteria, dataset name, index, reasoning for the first response, solution for the first response, split of LLM criteria, evaluation response, evaluation parsed, and score. The dataset is split into a training set (train) with 1000 examples and a size of 21577501 bytes. A default configuration is provided, along with the path to the training data files.
提供机构：
TAUR-dev
