File size: 1,260 Bytes
04bbd63
 
 
 
 
 
 
 
 
 
e174f9d
 
 
 
 
 
 
 
091bf1c
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
---
language:
- grc
tags:
- ELECTRA
- TensorFlow
---



An ELECTRA-small model for Ancient Greek, trained on texts from Homer up until the 4th century AD from the literary [GLAUx](https://github.com/alekkeersmaekers/glaux) corpus and the [DukeNLP](https://github.com/alekkeersmaekers/duke-nlp) papyrus corpus.

The model has some design choices made to combat data sparsity:
* Its input should always be in Unicode NFD (so separate Unicode signs for diacritics).
* All grave accents should be replaced with acute accents (καί, not καὶ).
* When a word contains two accents, the second one should be removed (εἶπε μοι, not εἶπέ μοι).

If you use it in conjunction with [glaux-nlp](https://github.com/alekkeersmaekers/glaux-nlp), you can pass the tokenized sentence to normalize_tokens from tokenization.Tokenization, using normalization_rule=greek_glaux, which will do all these normalizations for you. 

## Citation

```bibtex
@misc{mercelis_electra-grc_2022,
	title = {electra-grc},
	url = {https://huggingface.co/mercelisw/electra-grc},
	abstract = {An ELECTRA-small model for Ancient Greek, trained on texts from Homer up until the 4th century AD.},
	author = {Mercelis, Wouter and Keersmaekers, Alek},
	year = {2022},
}
```