This is a version of LaBSE-en-ru-myv-v1, fine-tuned for about 150K steps on the myv_ru_2022 dataset, in this notebook.
It demonstrates slighly better results than the v1 model, both on bitext mining and on the MLM task.
- Downloads last month
- 7
This is a version of LaBSE-en-ru-myv-v1, fine-tuned for about 150K steps on the myv_ru_2022 dataset, in this notebook.
It demonstrates slighly better results than the v1 model, both on bitext mining and on the MLM task.