Helsinki-NLP/opus-mt_tiny_nld-eng

OPUS-MT-tiny-nld-eng

Distilled model from a Tatoeba-MT Teacher: OPUS-MT-models/nl-en/opus-2019-12-05, 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_nld-eng"
>>> tokenizer = MarianTokenizer.from_pretrained(model_name)
>>> model = MarianMTModel.from_pretrained(model_name)
>>> tok = tokenizer("Hallo, hoe gaat het?", 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+	29.4	58.2	0.8329
Bouquet	52.8	64.5	0.875

Student

testset	BLEU	chr-F	COMET
Flores+	26.7	71.1	0.8886
Bouquet	49.3	68.5	0.8707