presencesw/multinli_contradiction

---
dataset_info:
features:
- name: gold_label
dtype: string
- name: anchor
dtype: string
- name: positive
dtype: string
- name: negative
dtype: string
splits:
- name: train
num_bytes: 69919546
num_examples: 274829
- name: dev_matched
num_bytes: 1948526
num_examples: 9815
- name: dev_mismatched
num_bytes: 2064087
num_examples: 9832
download_size: 30472261
dataset_size: 73932159
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: dev_matched
path: data/dev_matched-*
- split: dev_mismatched
path: data/dev_mismatched-*
---

The dataset includes four main features: gold_label (string type), anchor (string type), positive (string type), and negative (string type). It is divided into three parts: the training set (train) contains 274829 samples, the development matched set (dev_matched) contains 9815 samples, and the development mismatched set (dev_mismatched) contains 9832 samples. The download size of the dataset is 30472261 bytes, and the total size is 73932159 bytes. The dataset configuration is default, with data file paths corresponding to the training set, development matched set, and development mismatched set.

原始信息汇总

数据集概述

特征信息

• gold_label: 数据类型为字符串。
• anchor: 数据类型为字符串。
• positive: 数据类型为字符串。
• negative: 数据类型为字符串。

数据分割

• train: 包含274829个样本，占用69919546字节。
• dev_matched: 包含9815个样本，占用1948526字节。
• dev_mismatched: 包含9832个样本，占用2064087字节。

数据集大小

• 下载大小: 30472261字节。
• 数据集大小: 73932159字节。

配置信息

• default 配置包含以下数据文件：
  • train: 路径为 data/train-*
  • dev_matched: 路径为 data/dev_matched-*
  • dev_mismatched: 路径为 data/dev_mismatched-*