The first Old English to Modern English dataset for language translation!

The dataset is structured as `source_language, target_language` pairs. For more information on how the dataset was extracted, structured, annotated, parsed, etc, you can check [this][1] introductory notebook where the data collection methodology is explained and an LLM Gemma-2 model is finetuned on it.

Next you can find the original sources of these texts:
- The raw translations come from the excellent work of [Dr. Ophelia Hostetter](https://hostetter.camden.rutgers.edu/) and can be found in the [Old English Poetry Project](https://oldenglishpoetry.camden.rutgers.edu/). 
- The raw old english texts can be found in [The Complete Corpus of Anglo-Saxon Poetry](https://sacred-texts.com/neu/ascp/).

These texts translations encompass almost 79% of all extant Old English poetry (that’s 23,662 lines out of about 30,000 extant lines).

[1]: https://www.kaggle.com/code/alejopaullier/theoden-the-old-english-gemma

---
license: apache-2.0
task_categories:
- translation
- text-generation
language:
- en
size_categories:
- 1K<n<10K
---