Helsinki-NLP/opus-mt_tiny_eng-spa

OPUS-MT-tiny-eng-spa

Distilled model from a Tatoeba-MT Teacher: Tatoeba-MT-models/eng-spa/opusTCv20210807+bt_transformer-big_2022-03-13, which has been trained on the Tatoeba dataset.

We used the OpusDistillery to train new a new student with the tiny architecture, with a regular transformer decoder. For training data, we used Tatoeba. The configuration file fed into OpusDistillery can be found here.

How to run

from transformers import MarianMTModel, MarianTokenizer
model_name = "Helsinki-NLP/opus-mt_tiny_eng-spa"
tokenizer = MarianTokenizer.from_pretrained(model_name)
model = MarianMTModel.from_pretrained(model_name)
tok = tokenizer("The area is also home to species of animals and birds with a wide variety.", return_tensors="pt").input_ids
output = model.generate(tok)[0]
tokenizer.decode(output, skip_special_tokens=True)

Benchmarks

Teacher

testset	BLEU	chr-F	COMET
Flores+	28.5	56.4	0.8546
Bouquet	41.8	64.9	0.8787

Student

testset	BLEU	chr-F	COMET
Flores+	25.1	54.0	0.8280
Bouquet	36.8	61.7	0.8491