Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,12 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
---
|
| 4 |
+
**Acervus** is an extensively curated and cleaned Latin language dataset spanning all orthographies and periods.
|
| 5 |
+
|
| 6 |
+
It contains over **21 billion high quality** Latin tokens from a wide array of sources.
|
| 7 |
+
|
| 8 |
+
This includes nearly **1 million novel samples recovered from poorly digitized manuscripts**.
|
| 9 |
+
|
| 10 |
+
It is predominantly free of the OCR artifacts and errors. As of January 2026, it is likely the largest publicly available corpus of clean Latin text.
|
| 11 |
+
|
| 12 |
+
Samples are marked according to provenance and inferred quality. *Organic* samples were sourced primarily manually from a large number of established public Latin repositories. *Corrected* samples are derived from damaged examples available in LatinPD, refined with the aid of [Emendator](https://huggingface.co/aimgo/Emendator), an OCR correction model trained expressly for Latin. Samples obtained from CommonCrawl, CommonCorpus, LatinPD, FineWeb2, and FinePDF are similarly marked.
|