quickmt-ar-en Neural Machine Translation Model

quickmt-ar-en is a reasonably fast and reasonably accurate neural machine translation model for translation from ar into en.

Model Information

• Trained using eole
• 185M parameter transformer 'big' with 8 encoder layers and 2 decoder layers
• 20k sentencepiece vocabularies
• Exported for fast inference to CTranslate2 format
• Training data: https://huggingface.co/datasets/quickmt/quickmt-train.ar-en/tree/main

See the eole model configuration in this repository for further details and the eole-model for the raw eole (pytorch) model.

Usage with quickmt

You must install the Nvidia cuda toolkit first, if you want to do GPU inference.

Next, install the quickmt python library and download the model:

git clone https://github.com/quickmt/quickmt.git
pip install ./quickmt/

quickmt-model-download quickmt/quickmt-ar-en ./quickmt-ar-en

Finally use the model in python:

from quickmt import Translator

# Auto-detects GPU, set to "cpu" to force CPU inference
t = Translator("./quickmt-ar-en/", device="auto")

# Translate - set beam size to 5 for higher quality (but slower speed)
sample_text = 'نبه الدكتور إيهود أور -أستاذ الطب في جامعة دالهوزي في هاليفاكس، نوفا سكوتيا ورئيس الشعبة الطبية والعلمية في الجمعية الكندية للسكري- إلى أن البحث لا يزال في أيامه الأولى.'
t(sample_text, beam_size=5)

> 'Dr. Ehud Orr, professor of medicine at Dalhousie University in Halifax, Nova Scotia and head of the medical and scientific division of the Canadian Diabetes Association, warned that the research is still in its early days.'

# Get alternative translations by sampling
# You can pass any cTranslate2 `translate_batch` arguments
t([sample_text], sampling_temperature=1.2, beam_size=1, sampling_topk=50, sampling_topp=0.9)

> 'Professor of Medicine at Dalhousie University in Halifax, Nova Scotia and chairman of the Medical and Scientific Division at the Canadian Diabetes Society, cautioned that the research was still in its early days.'

The model is in ctranslate2 format, and the tokenizers are sentencepiece, so you can use ctranslate2 directly instead of through quickmt. It is also possible to get this model to work with e.g. LibreTranslate which also uses ctranslate2 and sentencepiece.

Metrics

bleu and chrf2 are calculated with sacrebleu on the Flores200 devtest test set ("ara_Arab"->"eng_Latn"). comet22 with the comet library and the default model. "Time (s)" is the time in seconds to translate (using ctranslate2) the flores-devtest dataset (1012 sentences) on an RTX 4070s GPU with batch size 32 (faster speed is possible using a large batch size).

	bleu	chrf2	comet22	Time (s)
quickmt/quickmt-ar-en	42.79	66.98	87.4	0.88
Helsink-NLP/opus-mt-ar-en	34.22	61.26	84.5	3.78
facebook/nllb-200-distilled-600M	39.13	64.14	86.22	21.58
facebook/nllb-200-distilled-1.3B	42.29	66.55	87.55	37.7
facebook/m2m100_418M	29.41	57.68	82.21	18.5
facebook/m2m100_1.2B	29.77	56.7	80.77	36.23

quickmt-ar-en is the fastest and highest quality.