en-vi / README.md
andreacavallo's picture
Create README.md
d2d3cd6
|
raw
history blame
774 Bytes
metadata
language:
  - en
  - vi
tags:
  - translation
license: apache-2.0
datasets:
  - ALT
metrics:
  - sacrebleu

This is a finetuning of a MarianMT pretrained on English-Chinese. The target language pair is English-Vietnamese. The first phase of training (mixed) is performed on a dataset containing both English-Chinese and English-Vietnamese sentences. The second phase of training (pure) is performed on a dataset containing only English-Vietnamese sentences.

Training results

MIXED

Epoch Bleu
1.0 26.2407
2.0 32.6016
3.0 35.4060
4.0 36.6737
5.0 37.3774

PURE

Epoch Bleu
1.0 37.3169
2.0 37.4407
3.0 37.6696
4.0 37.8765
5.0 38.0105