---
language: sem
tags:
- translation

license: apache-2.0
---

### sem-sem

* source languages: sem
* target languages: sem
*  OPUS readme: [sem-sem](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/sem-sem/README.md)

*  dataset: opus
* model: transformer
* source language(s): apc ara arq arz heb mlt
* target language(s): apc ara arq arz heb mlt
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* a sentence initial language token is required in the form of `>>id<<` (id = valid target language ID)
* download original weights: [opus-2020-07-27.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/sem-sem/opus-2020-07-27.zip)
* test set translations: [opus-2020-07-27.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/sem-sem/opus-2020-07-27.test.txt)
* test set scores: [opus-2020-07-27.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/sem-sem/opus-2020-07-27.eval.txt)

## Benchmarks

| testset               | BLEU  | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.ara-ara.ara.ara 	| 4.2 	| 0.200 |
| Tatoeba-test.ara-heb.ara.heb 	| 34.0 	| 0.542 |
| Tatoeba-test.ara-mlt.ara.mlt 	| 16.6 	| 0.513 |
| Tatoeba-test.heb-ara.heb.ara 	| 18.8 	| 0.477 |
| Tatoeba-test.mlt-ara.mlt.ara 	| 20.7 	| 0.388 |
| Tatoeba-test.multi.multi 	| 27.1 	| 0.507 |