README.md

metadata

license:
  - cc-by-4.0
  - cc-by-sa-4.0
  - cc0-1.0
language:
  - de
  - en
  - es
  - fr
  - it
  - ru
  - zh
task_categories:
  - text-generation
size_categories:
  - n<1K
num_rows: 10
num_languages: 7
train_count: 6
validation_count: 4

Multilingual Short Stories

A small multilingual short story catalog curated from the source workbook file1.xlsx. Records with empty or very short text (fewer than 50 whitespace-delimited words) were removed, language values were standardized to two-letter ISO 639-1 codes, and a split column was added (train when word_count >= 200, otherwise validation).

Dataset Summary

• Total rows: 10
• Number of unique languages: 7
• Train rows: 6
• Validation rows: 4

Languages

Language	Rows
de	1
en	4
es	1
fr	1
it	1
ru	1
zh	1

Splits

Split	Rows
train	6
validation	4

Data Fields

• id: unique identifier of the story
• title: title of the story
• language: two-letter ISO 639-1 language code
• source: origin of the record
• license: license of the story text
• word_count: reported word count
• text: full story text
• split: dataset split (train or validation)

Licenses

cc-by-4.0, cc-by-sa-4.0, cc0-1.0

Intended Uses

This dataset is intended for text generation and language modeling experiments across multiple languages.