NyanNyanovich/nyan_clusters

---
dataset_info:
features:
- name: docs
sequence: string
- name: message_id
dtype: float64
- name: create_time
dtype: float64
- name: annotation_doc
dtype: string
- name: first_doc
dtype: string
- name: hash
dtype: string
- name: is_important
dtype: bool
- name: clid
dtype: float64
- name: message
struct:
- name: from_discussion
dtype: bool
- name: issue
dtype: string
- name: message_id
dtype: int64
- name: messages
list:
- name: from_discussion
dtype: bool
- name: issue
dtype: string
- name: message_id
dtype: int64
splits:
- name: train
num_bytes: 21796882
num_examples: 41296
download_size: 6836100
dataset_size: 21796882
license: cc-by-4.0
language:
- ru
pretty_name: Nyan Clusters
size_categories:
- 10K<n<100K
---

# Nyan clusters

Clusters of documents formed in [НЯН](https://t.me/nyannews) Telegram channel from March 2022 to December 2023. You can use the documents dataset to get texts.

## Usage
```bash
pip3 install datasets
```

```python
from datasets import load_dataset
for row in load_dataset("NyanNyanovich/nyan_clusters", split="train", streaming=True):
print(row)
break
```

## Other datasets
* Documents: https://huggingface.co/datasets/NyanNyanovich/nyan_documents
* Clusters (this dataset): https://huggingface.co/datasets/NyanNyanovich/nyan_clusters

原始信息汇总

数据集概述

数据集信息

• 特征列表:

  • docs: 字符串序列
  • message_id: 浮点数 (float64)
  • create_time: 浮点数 (float64)
  • annotation_doc: 字符串
  • first_doc: 字符串
  • hash: 字符串
  • is_important: 布尔值 (bool)
  • clid: 浮点数 (float64)
  • message: 结构体
    • from_discussion: 布尔值 (bool)
    • issue: 字符串
    • message_id: 整数 (int64)
  • messages: 列表
    • from_discussion: 布尔值 (bool)
    • issue: 字符串
    • message_id: 整数 (int64)

• 数据分割:

  • train:
    • 字节数: 21796882
    • 样本数: 41296

• 下载大小: 6836100 字节

• 数据集大小: 21796882 字节

• 许可: cc-by-4.0

• 语言: 俄语 (ru)

• 名称: Nyan Clusters

• 大小类别: 10K<n<100K