Use different language to compare similarities of sentences

pdrm83 / sent2vec

How to encode sentences in a high-dimensional vector space, a.k.a., sentence embedding.

MIT License

132 stars 12 forks source link

Use different language to compare similarities of sentences #5

Open simongehlhar opened 3 years ago

simongehlhar commented 3 years ago

Is it possible to compare sentences in German by maybe using a different model? How would one use a different model within your framework? For example: 'https://huggingface.co/transformers/pretrained_models.html: bert-base-german-cased'

Regards

pdrm83 commented 3 years ago

I didn't test it. It must work; though. Fundamentally, I have seen no difference. Note that I used Transformers lib in this package. The main point of this package is the rapid and flexible prototyping for sentence vectorization. As I explained in the repo description, I use standard libs. Please feel free to contribute to this open-source package.

almarengo commented 2 years ago

You can find this useful to solve your issue. Please refer to #9

I hope it helps.