legal_t5_small_trans_sv_fr model
Model on translating legal text from Swedish to French. It was first released in
this repository. This model is trained on three parallel corpus from jrc-acquis, europarl and dcep.
Model description
legal_t5_small_trans_sv_fr is based on the t5-small model and was trained on a large corpus of parallel text. This is a smaller model, which scales the baseline model of t5 down by using dmodel = 512, dff = 2,048, 8-headed attention, and only 6 layers each in the encoder and decoder. This variant has about 60 million parameters.
Intended uses & limitations
The model could be used for translation of legal texts from Swedish to French.
How to use
Here is how to use this model to translate legal text from Swedish to French in PyTorch:
1from transformers import AutoTokenizer, AutoModelWithLMHead, TranslationPipeline
2
3pipeline = TranslationPipeline(
4model=AutoModelWithLMHead.from_pretrained("SEBIS/legal_t5_small_trans_sv_fr"),
5tokenizer=AutoTokenizer.from_pretrained(pretrained_model_name_or_path = "SEBIS/legal_t5_small_trans_sv_fr", do_lower_case=False,
6 skip_special_tokens=True),
7 device=0
8)
9
10sv_text = "Kunden måste ha rätt att avsäga sig information i skriftlig form."
11
12pipeline([sv_text], max_length=512)
Training data
The legal_t5_small_trans_sv_fr model was trained on
JRC-ACQUIS,
EUROPARL, and
DCEP dataset consisting of 5 Million parallel texts.
Training procedure
The model was trained on a single TPU Pod V3-8 for 250K steps in total, using sequence length 512 (batch size 4096). It has a total of approximately 220M parameters and was trained using the encoder-decoder architecture. The optimizer used is AdaFactor with inverse square root learning rate schedule for pre-training.
Preprocessing
An unigram model trained with 88M lines of text from the parallel corpus (of all possible language pairs) to get the vocabulary (with byte pair encoding), which is used with this model.
Pretraining
Evaluation results
When the model is used for translation test dataset, achieves the following results:
Test results :
| Model | BLEU score |
|---|
| legal_t5_small_trans_sv_fr | 47.623 |
BibTeX entry and citation info
