---
format: "aidr-story-markdown/v1"
id: "6acfc421b69e1daf0ccfe81d84934a3d82c6569b942fcba4de458bf606f1e34a"
canonical_url: "https://aidr.today/6acfc421?lang=en"
title: "Wieszcz-XIX: A 3.1-Billion-Word Corpus of Pre-1918 Polish and Temporally Bounded Language Models Trained From Scratch"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-09T04:00:00.000Z"
category: "Data"
topics: ["llm","data-engineering"]
source_urls: ["https://arxiv.org/abs/2610.10592"]
summary: "Wieszcz-XIX is a corpus of 6.75 billion tokens (about 3.1 billion words) in 294,369 documents of Polish published from 1800 to 1918. The corpus was assembled from Wolne Lektury and the Internet Archive using a pipeline that filters, deduplicates, and audits for post-1918 leakage."
---

# Wieszcz\-XIX: A 3\.1\-Billion\-Word Corpus of Pre\-1918 Polish and Temporally Bounded Language Models Trained From Scratch

> [Open the canonical story](<https://aidr.today/6acfc421?lang=en>)

**Published:** 2026-10-09T04:00:00.000Z
**Category:** Data
**Topics:** llm, data\-engineering

## Summary

Wieszcz\-XIX is a corpus of 6\.75 billion tokens \(about 3\.1 billion words\) in 294,369 documents of Polish published from 1800 to 1918\. The corpus was assembled from Wolne Lektury and the Internet Archive using a pipeline that filters, deduplicates, and audits for post\-1918 leakage\.

## Sources

- [Story source](<https://arxiv.org/abs/2610.10592>)