latam-gpt/gretel-safety-alignment-es-v1

该数据集是gretelai/gretel-safety-alignment-en-v1数据集的西班牙语翻译版本，包含了关于歧视、信息风险、恶意使用、社会风险和系统风险五个方面的数据，每个方面都有训练集、测试集和验证集。数据集中的字段包括提示（prompt）、响应（response）和安全响应（safe_response），翻译工作主要由gpt4o-mini模型完成，部分由mistral-small-3.2-24b-instruct模型完成。此外，原数据集中包含模型拒绝的行已经被过滤掉。

This dataset is the Spanish translation version of the gretelai/gretel-safety-alignment-en-v1 dataset, containing data on discrimination, information hazards, malicious use, societal risks, and system risks, each with training, test, and validation sets. The dataset includes fields for prompt, response, and safe_response, with translation done mainly by the gpt4o-mini model and some by the mistral-small-3.2-24b-instruct model. Additionally, rows containing model refusals in the original dataset have been filtered out.