Add jina-embeddings-v2-base-es

#9

I just copy the files from https://huggingface.co/jinaai/jina-embeddings-v2-base-es, there was no vocab.txt in that repo just vocab.json, I've converted it from json to txt, I'm not sure it will work, if needed I can upload the vocab.json file too.

hcentelles changed pull request title from Adding jina-embeddings-v2-base-es/config.json to Add jina-embeddings-v2-base-es
Typesense org

@hcentelles did you confirm the model is working on your local with Typesense?

Nope :| I'm not running Typesense local, just cloud. Sorry, that's why I submitted a request to include the model before the PR.

Typesense org

I will try and will inform you.

Typesense org

@hcentelles The model you want to add uses RoBERTa tokenizer, which is a different tokenizer than original BERT tokenizer. We do not support it yet, do you have any other alternatives to this model for now?

No, I don't. Thank you very much for trying it. I can close the PR if needed.

Typesense org

Ok I will close the PR. By the way, I can recommend giving a try to paraphrase-multilingual-mpnet-base-v2.

ozanarmagan changed pull request status to closed

Sign up or log in to comment