Translation
LiteRT
Safetensors
English
French
marian
odegiber's picture
Update README.md
4447b69 verified
metadata
license: apache-2.0
datasets:
  - Helsinki-NLP/tatoeba
  - openlanguagedata/flores_plus
  - facebook/bouquet
language:
  - en
  - fr
metrics:
  - bleu
  - comet
  - chrf
pipeline_tag: translation

OPUS-MT-tiny-fra-eng

Distilled model from a Tatoeba-MT Teacher: OPUS-MT-models/fr-en/opus-2020-02-26, 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_fra-eng"
tokenizer = MarianTokenizer.from_pretrained(model_name)
model = MarianMTModel.from_pretrained(model_name)
tok = tokenizer("Les efforts visant à trouver le lieu de l’accident sont restreints par des intempéries et le terrain accidenté.", 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+ 41.8 66.9 0.8689
Bouquet 43.5 64.5 0.875

Student

testset BLEU chr-F COMET
Flores+ 40.4 65.9 0.8734
Bouquet 39.9 61.9 0.8551