buseskorkmaz/hiring_w_q_context_256_filtered_biasinbios_occ_q_val_w_noise

--- dataset_info: features: - name: text dtype: string - name: CommentTime dtype: string - name: CommentAuthor dtype: string - name: ParentTitle dtype: string - name: clean_text dtype: string - name: location dtype: string - name: technologies dtype: string - name: embedding sequence: float64 - name: company dtype: string - name: title dtype: string - name: prompt dtype: string - name: q_value dtype: float64 - name: biasinbios_occupations dtype: string - name: evaluated_text dtype: string - name: sr_female dtype: float64 - name: sr_male dtype: float64 - name: gender_distance dtype: float64 - name: ir_female dtype: float64 - name: ir_male dtype: float64 - name: q_val dtype: float64 - name: male_profession_percentages struct: - name: accountant dtype: float64 - name: attorney dtype: float64 - name: journalist dtype: float64 - name: professor dtype: float64 - name: software_engineer dtype: float64 - name: female_profession_percentages dtype: string - name: male_selected_profession_percentages dtype: string - name: female_selected_profession_percentages dtype: string - name: matched_professions sequence: string - name: noise dtype: string splits: - name: train num_bytes: 70690588 num_examples: 5635 download_size: 45267187 dataset_size: 70690588 configs: - config_name: default data_files: - split: train path: data/train-* ---The dataset includes multiple features such as text, comment time, comment author, parent title, clean text, location, technologies, embedding vectors, company, title, prompt, q-value, bias in occupations, evaluated text, gender ratios, gender distance, profession percentages, etc. Each feature has its specific data type, such as string, float64, etc. The dataset is divided into a training set, containing 5635 samples, with a total size of 70690588 bytes.

原始信息汇总

数据集概述

数据集特征

• text: 字符串类型
• CommentTime: 字符串类型
• CommentAuthor: 字符串类型
• ParentTitle: 字符串类型
• clean_text: 字符串类型
• location: 字符串类型
• technologies: 字符串类型
• embedding: 序列类型，浮点数64位
• company: 字符串类型
• title: 字符串类型
• prompt: 字符串类型
• q_value: 浮点数类型
• biasinbios_occupations: 字符串类型
• evaluated_text: 字符串类型
• sr_female: 浮点数类型
• sr_male: 浮点数类型
• gender_distance: 浮点数类型
• ir_female: 浮点数类型
• ir_male: 浮点数类型
• q_val: 浮点数类型
• male_profession_percentages: 结构类型
  • accountant: 浮点数类型
  • attorney: 浮点数类型
  • journalist: 浮点数类型
  • professor: 浮点数类型
  • software_engineer: 浮点数类型
• female_profession_percentages: 字符串类型
• male_selected_profession_percentages: 字符串类型
• female_selected_profession_percentages: 字符串类型
• matched_professions: 序列类型，字符串
• noise: 字符串类型

数据集分割

• train:
  • 数据大小: 70690588字节
  • 示例数量: 5635

数据集大小

• 下载大小: 45267187字节
• 数据集总大小: 70690588字节

配置

• config_name: default
  • data_files:
    • split: train
      • path: data/train-*