Dataset Viewer
The dataset viewer is not available for this subset.
Job has been terminated due to a temporary spike in resource usage and may be restarted later.
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
wikisource-journals
Russian literary texts from ru.wikisource.org: prose, poetry, journalism, criticism.
- split:
train, 24,639 rows - parquet (zstd) in
data/, 4 shards, ~705 MB - total length: 1.29B chars; median 9,906 chars; max 2.4M chars
Schema
| field | type | notes |
|---|---|---|
id |
string | sha1 of source path, first 16 hex chars |
title |
string | Wikisource page title |
text |
string | cleaned full text |
url |
string? | page on ru.wikisource.org; null for 636 rows without metadata |
author |
string? | null for 361 rows |
translator |
string? | parsed from Title (Author; Translator) filenames |
category |
string | author (826) / journal (4,103) / translation (12,125) / thematic (7,585) |
publication_date |
string? | filled for 54 rows only |
year |
int32? | year parsed from filename; filled for 5,809 rows |
orthography |
string? | дореволюционная / современная |
n_chars |
int32 | length of text |
sha1 |
string | sha1 of source file |
Load
from datasets import load_dataset
dataset = load_dataset("parquet", data_files="data/train-*.parquet", split="train")
License
Mostly public domain (published before 1917).
- Downloads last month
- 36