alsunyu/CONSCENDI-guardrail-benchmark

Name: alsunyu/CONSCENDI-guardrail-benchmark
Creator: alsunyu
Published: 2024-03-29 17:39:13
License: 暂无描述

Hugging Face2024-03-29 更新2024-05-25 收录

下载链接：

https://hf-mirror.com/datasets/alsunyu/CONSCENDI-guardrail-benchmark

下载链接

链接失效反馈

官方服务：

资源简介：

--- dataset_info: features: - name: conversation dtype: string - name: rule_num dtype: string - name: rule dtype: string - name: scenario_num dtype: float64 - name: scenario dtype: string - name: english_num dtype: float64 - name: prompt dtype: string - name: time_elapsed dtype: float64 - name: cost dtype: float64 - name: prompt_tokens dtype: float64 - name: completion_tokens dtype: float64 - name: generation dtype: string - name: convo_length dtype: float64 - name: completion dtype: string - name: spliced_section dtype: string - name: rule_scenario dtype: string splits: - name: bus_test_id num_bytes: 386861 num_examples: 351 - name: bus_test_ood num_bytes: 288372 num_examples: 255 - name: bus_train num_bytes: 1020209 num_examples: 946 - name: flight_test_id num_bytes: 457393 num_examples: 347 - name: flight_test_ood num_bytes: 390962 num_examples: 302 - name: flight_train num_bytes: 1255875 num_examples: 937 - name: restaurant_test_id num_bytes: 426197 num_examples: 334 - name: restaurant_test_ood num_bytes: 324046 num_examples: 298 - name: restaurant_train num_bytes: 1162826 num_examples: 901 download_size: 1731265 dataset_size: 5712741 license: mit task_categories: - text-classification language: - en tags: - legal pretty_name: CONSCENDI Rule-based Guardrail Benchmark size_categories: - 1K<n<10K --- # Dataset Card for "CONSCENDI-guardrail-benchmark" [More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)

提供机构：

alsunyu

原始信息汇总

数据集概述

数据集名称

CONSCENDI Rule-based Guardrail Benchmark

数据集特征

conversation: 字符串
rule_num: 字符串
rule: 字符串
scenario_num: 浮点数
scenario: 字符串
english_num: 浮点数
prompt: 字符串
time_elapsed: 浮点数
cost: 浮点数
prompt_tokens: 浮点数
completion_tokens: 浮点数
generation: 字符串
convo_length: 浮点数
completion: 字符串
spliced_section: 字符串
rule_scenario: 字符串

数据集分割

bus_test_id: 351个样本，386861字节
bus_test_ood: 255个样本，288372字节
bus_train: 946个样本，1020209字节
flight_test_id: 347个样本，457393字节
flight_test_ood: 302个样本，390962字节
flight_train: 937个样本，1255875字节
restaurant_test_id: 334个样本，426197字节
restaurant_test_ood: 298个样本，324046字节
restaurant_train: 901个样本，1162826字节

数据集大小

下载大小: 1731265字节
数据集大小: 5712741字节

许可证

MIT

任务类别

文本分类

语言

英语

大小类别

1K<n<10K

5,000+

优质数据集

54 个

任务类型

进入经典数据集