angelahzyuan/data-mistral-7b-instruct-iter6

数据集data-mistral-7b-instruct-iter6包含训练和测试两个分割，主要用于指令生成任务。数据集的特征包括prompt_id、prompt、chosen、rejected、all_generated_responses和all_rm_scores。chosen和rejected字段是列表类型，包含content和role两个子字段，分别表示生成的内容和角色。all_generated_responses和all_rm_scores是序列类型，分别存储所有生成的响应和对应的评分。训练集包含9998个样本，测试集包含500个样本。

The dataset data-mistral-7b-instruct-iter6 includes training and test splits, primarily used for instruction generation tasks. The features of the dataset include prompt_id, prompt, chosen, rejected, all_generated_responses, and all_rm_scores. The chosen and rejected fields are of list type, containing subfields content and role, representing the generated content and role respectively. all_generated_responses and all_rm_scores are of sequence type, storing all generated responses and their corresponding scores. The training set contains 9998 examples, and the test set contains 500 examples.