metadata
language:
- ar
- en
tags:
- translation
license: cc-by-4.0
Translation model for ar-en HPLT v1.0
This repository contains the model weights for translation models trained with Marian for HPLT project. For usage instructions, evaluation scripts, and inference scripts, please refer to the HPLT-MT-Models v1.0 GitHub repository.
- Source language: ar
- Target language: en
- Dataset: HPLT only
- Model: transformer-base
- Tokenizer: SentencePiece (Unigram)
- Cleaning: We use OpusCleaner for cleaning the corpus. Details about rules used can be found in the filter files in Github
To run inference with Marian, refer to the Inference/Decoding/Translation section of our GitHub repository.
Benchmarks
| testset | BLEU | chr-F | COMET-22 |
|---|---|---|---|
| flores200 | 35.0 | 58.5 | 0.8396 |
| ntrex | 28.6 | 54.6 | 0.8194 |